The Impact Of Dividend Policy

Published Date: 02 Nov 2017

The impact of dividend policy on the stock price of a firm is still a controversial issue in the field of finance over last few decades. Many researchers and economists have engaged in analyzing and testing company dividend policy since many years ago. However, there are still continuing debates on this issue until today.

Dividend is a portion of a firm’s current or retained earnings distributed to its shareholders. It can be in the form of cash or share of the firm. Dividend policy is commonly defined as a policy implemented by the firm to decide how much dividend it will pay out to its shareholders. The management has to decide whether to pay out or retain the profits once the firm makes profit. The firm may implement a dividend policy which may impact the investors and future prospects of the firm in the financial markets once the firm decides to pay out dividends to its shareholders. The decision depends on the current situation of the firm and also the preferences of investors. The main objective of the management is to maximize the shareholders’ wealth. Therefore, this objective can be achieved by paying an appropriate amount of dividend to the shareholders on their investments. However, the impact of dividend decisions on shareholders’ wealth is still a debatable issue.

In general, the market price of a firm’s stock represents the shareholders’ wealth which is important for the firm’s investment, financing and dividend decisions. In a sense, the dividend decision is essentially important for the achievement of the firm’s objectives and efficient performance because the role of the finances is increasingly significant in the firm’s growth strategy. Therefore, management of the company has to determine an appropriate dividend policy in order to increase the value of the firm. There is an argument that a decrease in the dividend payment will reduce the market share price of the firm.

There are several types of dividend policy such as stable dividend policy, constant dividend policy and residual dividend policy. Stable dividend policy offers a stable dividend payout ratio. It has a positive impact on the firm’s share price. It includes constant dividend per share and constant payout ratio. The management will not simply decrease the dividend until the company is incapable to support the present dividend. On the other hand, residual dividend policy states that the dividend should be paid only out of left over earnings. This is because the investors prefer the firm to retain the money for further investment if the return on the firm’s investment exceeds that of the individual investor investment of a similar risk. The company can also save on issue cost such as legal fee, underwriting fee and underpricing.

Besides that, found that companies have long run target dividend payout ratio which can only be achieved in the long run. Management also focuses more on dividend changes rather than on absolute value. Besides that, he also found that managers smooth dividend. Therefore, temporary changes in earnings are unlikely to lead to changes in dividend as managers do not like erratic dividend payment due to uncertainty. Furthermore, managers are reluctant to make dividend changes that might have to be reversed as this will lead to uncertainty among shareholders.

On the other hand, argue that dividend is irrelevant and does not affect the firm’s value under perfect capital market. The reason is that there is no agency cost, asymmetric information, tax, transaction cost and flotation cost in perfect capital market.

Many past studies state that dividend is a signaling tool. This is known as dividend signaling theory. It indicates that high dividend payment provides a positive signal about the current and future prosperity of a company to outsiders. This is in line with the efficient market hypothesis where all information available to investors is quickly and accurately incorporated into the share price. However, some researchers found no evidence of that. Therefore, another dividend issue still remains unsolved.

Moreover, dividend can be used to reduce agency problem. Agency theory deals with the potential conflict of interest between shareholders and management. Dividend payment will reduce the amount of free cash flow within the company. Therefore, it reduces the ability of manager to waste shareholders’ funds on questionable acquisition. As a consequence, dividend payment is relevant as it reduces agency cost and indirectly increases the firm value.

In Malaysia, firms are free to decide how much they will pay out in dividend to their shareholders for a specified business year in accordance with Companies Act 1965. Shareholders prefer cash dividend and also growth in earning per share (EPS). Firms should implement optimal dividend policy that maximizes the market share price of the firm and therefore improves economic growth. The study highlights the relationship between dividend payout and market share price of a firm. It also examine whether there is impact of dividend policy on shareholders’ wealth.

Problem Statement

The effect of dividend policy on share price is still a controversial topic for many years. Dividend refers to a portion of a firm’s current or retained earnings distributed to its shareholders. Whether it has impact on the market share price of a firm is still questionable. None of researchers and financial analysts can actually determine the best dividend policy. Some researchers believe that dividend payment will increase the firm value. However, some believe that dividend payment will decrease the firm value instead. Furthermore, some researchers also emphasize that dividend payment has no impact on the firm value. For instance, states that dividend payment and share price are irrelevant under perfect capital market where there is no tax, transaction cost and asymmetric information.

Furthermore, leftist states that a firm will pay less or no dividend if dividend is taxed more heavily than capital gain. The money should be retained within the firm to finance new investment for future capital gain. Conversely, found that investors prefer high dividend policy as dividend in hand is less risky than capital gain. Besides that, Watts (1973) found that a company’s current and past dividends did have a little effect on the company’s future prospects. As a result, managers are difficult to determine which dividend policy is best fit to the company.

Why managers still intend to pay dividend to the company’s shareholders if dividend is irrelevant to the share price? If dividend does have an impact on the firm’s value, which dividend policy theory can best describe the relationship between dividend payout and share price? Therefore, this study wants to explain whether cash dividend payment and other stated factors (earnings per share, firm size and debt ratio) and share price are relevant in selected Malaysian companies.

Objectives

Objectives of this research are:

To examine the relationship between dividend payment and the share price of the selected companies in Malaysia.

To determine which dividend policy theory corresponds with Malaysian companies’ dividend phenomenon.

To identify whether other factors such as earnings per share, firm size and debt ratio play a role in the impact of share price of selected Malaysian companies.

Hypotheses

• Hypothesis 1:

H0: There is no significant statistical relationship between cash dividend and the market share price of companies in Malaysia.

H1: There is a significant statistical relationship between cash dividend and the market share price of companies in Malaysia.

• Hypothesis 2:

H0: There is no significant statistical relationship between earnings per share (EPS) and the market share price of companies in Malaysia.

H1: There is a significant statistical relationship between earnings per share (EPS) and the market share price of companies in Malaysia.

• Hypothesis 3:

H0: There is no significant statistical relationship between firm size and the market share price of companies in Malaysia.

H1: There is a significant statistical relationship between firm size and the market share price of companies in Malaysia.

• Hypothesis 4:

H0: There is no significant statistical relationship between debt ratio and the market share price of companies in Malaysia.

H1: There is a significant statistical relationship between debt ratio and the market share price of companies in Malaysia.

Research Design

The deductive approach is used to formulate hypotheses and measure them empirically. Quantitative approach relies on numerical evidence to form conclusions or to test hypothesis. Analytical designs are used in this research for the analysis of the relationship between dividend payments and market share price of Malaysian companies to achieve the objectives of this research. Explanatory method is also applied in this research to explain the relationship between the four chosen factors and market value of share of those selected Malaysian companies. Explanatory method is used as it could recognize and meet the purpose of explaining the relationship between the independent variables and the dependent variable.

Data collection

103 companies in Malaysia from several industries are chosen as the sample. The requirement is that companies should have paid cash dividend for the past five financial years (year 2008-year 2012). Secondary sources are collected from library, reference books and journals, corporation financial statements and audit reports which similarly conducted by other researchers. These secondary data are used for references and aid in research writing.

Resources

Library and its resources - reference books and articles in the library which are enrich with information assisting research findings.

Computer and its software - Laptop is used for data storage, writing purpose using Microsoft words and presentation purpose using Power Point. Statistical Package for Social Sciences (SPSS) program is used to analyze and test the hypotheses by calculating the means, standard deviations, correlation coefficient.

Internet - Internet is significant for online information gathering using search engines such as Google or EBSCO host and communication between group members via electronic-mail.

2.0 Literature Review

2.1 The Impact of Various Dividend Policies on Share Price

2.1.1 Irrelevant Dividend Policy

concluded that dividend policy has no impact on a company’s share price under perfect capital market, that is, with no agency cost, tax, asymmetric information, transaction cost and floatation cost. The original shareholders transfer value to new shareholders when the firm issues new shares to finance the dividend. They claimed that only the firm’s investment policy will affect its share price. As a result, dividend policy will not help managers to maximize shareholders’ wealth.

tested the relationship between share return and dividend yield based on 25 portfolios of ordinary shares listed on NYSE for the period from 1936 to 1966. The purpose was to determine the impact of dividend policy on market share price. They applied capital asset pricing model (CAPM) for testing the relationship between dividend yield and expected share return. Their results found that the coefficient of dividend yield was not significantly different from zero. Their research empirically supported the theory and concluded that the returns for investors who are holding low or high yield shares are the same.

However, concluded that the dividend policy is relevant since he tested the impact of dividend on the share price based on Australian data from the period of 1960 to 1969 by applying the cross-sectional regressions model. He stated that the no-retention assumption created by M&M does not make sense as it is impossible for a firm to distribute all its free cash flow to its shareholders and thus each firm will retain some profits for further investment. As a result, the dividend policy is said to be relevant because suboptimal policies could be chosen by managers to invest in positive NPV projects. Besides that, interviewed 603 listed US firms’ CFOs by designing questionnaire and did a survey about their behaviors. The survey proves that ninety percent of interviewers did not agree with M&M’s irrelevant dividend policy theory. It also showed that 87% of respondents agree with the interrelationship between dividend and the firm’s financing and investment policies.

2.1.2 Dividend Policy based on Bird-In-Hand Theory

conducted this study in 1959 and concluded that investors prefer dividend payments to capital gains as dividends are safer and they provide investors a secure income. He applied a two-year (1951 and 1954) sample data of four industries including steels, food, chemical and manufacturing tools industries. He then tested them by analyzing various regression models based on the data. The dividend hypothesis was examined by using linear regression model. The results proved that future capital gains received by investors will have higher fluctuation than dividends today. According to his study, higher dividend payout will cause investors to discount a higher dividend stream at a lower required rate of return. He also argued that a firm’s shares with high dividend payout normally have a lower risk. In a sense, less risky shares will normally have a higher price given that other factors of determining share price remain unchanged. As a result, this combined effect will then increase the firm’s share price and its value as the risk (ke) will be lower when the firm distributes dividend to shareholders.

reexamined Gordon model and corrected his regressions model equation by changing R to a three-year mean value centered on t-1. He tested the effect of retained profits and dividends on stock prices based on 255 listed US companies from 8 industries for the year of 1961 and 1962. His results showed that high growth industries prefer more retained profits than dividends whereas low growth industries prefer dividends to retained profits. As a result, he stated that the growth of company associated negatively with the dividend payout. His outcome was quite similar to early results

concluded that there is a positive relationship between dividend payout ratio and the volatility of the share price based on their research on 160 Pakistani companies for the period from year 1981 to year 2000 based on cross sectional regressions model. However, stated that investors of Nigerian companies prefer the company distribute the dividends to them based on their expectations. By using Pearson Product Moment correlation model, they claimed that there is no relationship between dividend payout, net profit and the company’s value as company distributes dividend to satisfy its shareholders regardless of its profit earned.

Moreover, interviewed NASDAQ companies’ managers to examine the Bird-in-Hand effect on market share price. Their results found that 17.2% of 186 responses prefer dividends to future capital gains, 28% of them remained neutral, and 54.9% of them rejected the Bird-in-Hand theory. They then concluded that their study did not empirically give evidence to support the bird-in-hand definition.

2.1.3 Dividend Policy based on Agency Theory

Agency theory deals with the potential conflict of interest between shareholders and management. Conflicts happen when managers ignore shareholders’ interests and focus only on their personal goals. As a result, shareholders may want to control managers’ behaviors through some policies. Monitoring managers will bring high costs to a company. Generally, shareholders may find it difficult to control managers’ actions as tasks and decisions made by managers are most likely difficult to be understood by shareholders. Therefore, a specialized outsider plays an important role for monitoring them. For instance, managers can be monitored effectively through external financing. In brief, dividend payment will decrease the amount of free cash flow within a company. Therefore, shareholders’ fund will not be simply wasted by managers on questionable acquisitions since the reduction of cash will negatively affect managers’ performance. Managers will then issue equity to finance the investments once they get approval from Security Commission. In a sense, Security Commission is aiding shareholders to monitor managers’ behaviors. As a result, dividend payment is said to be relevant since it helps lower agency cost and indirectly increases the firm value and its market share price.

Furthermore, found that dividend policy will decrease managers’ cash in hand and thus it leads them to borrow money from outsider to finance the firm’s investment. This can help shareholders to control managers’ actions through extra control of external borrower. He developed a model that underpins this theory, called the cost minimization model. The model combines the transaction costs that may be controlled by limiting the payout ratio, with the agency costs that may be controlled by raising the payout ratio. Indeed results from an Ordinary Least Squares cross sectional regression using 1981 data on 1000 US firms, support the theory put forward. Thus the model provides good fit and consequently has attracted the attention of subsequent studies.

found that it is not profitable for a company to perform merger as the price for merger is most likely higher than the market value. However, managers tend to purchase other companies for their desires to raise scope of control within the company. As a result, this agency cost can be eliminated by simply paying higher dividends in order to avoid over-investment by managers. Furthermore, modified Rozeff’s cost minimization model by adding a size variable. An OLSQ cross sectional regression is applied to 1984 data on 957 US firms, and the results provide support for the cost minimization model and show that firm size is an important explanatory variable.

observed that managers normally know more about the firm’s performance than its shareholders due to asymmetric information. However, managers may not be able to disclose information to the firm’s shareholders due to reasons like difficulties to prove the information or for the fear of being sued by Security Commission. Therefore, managers will use other method like paying higher dividend to provide positive signal to the company’s shareholders. As a result, the information gap between shareholders and managers will be smaller.

2.1.4 Dividend Policy based on Signaling Theorytested the market response to dividend announcements based on sample of 168 companies that pay dividend for the first time or continued distributing dividends after ten-year hiatus. They examined the average daily excess share returns ten days surrounding the notice of dividend initiation. Their results showed that positive 3.7 percent of excess return was found for the period of two-day announcement. Furthermore, they stated that there was a significant relationship between the importance of initial dividend and excess share returns on the day of announcement by applying the cross-sectional regression.

carried forward the results. In his model, it investigated the importance of dividend in two different ways. Firstly the relation whether the unexpected dividend changes were associated with positive future earning changes and secondly the relation between unexpected changes in share price and changes in annual dividends which convey priceless information to shareholders. The sample consisted of 310 firms. The results showed that relationship was positive but not very strong.

provided evidence for the statement that modification of dividend communicates information about companies’ values. They applied 255 dividend increase events and 51 dividend decrease events from year 1988 to year 1992 to test the impact of announcement on share price in US. In their model, they base their argument on the declining dividend response coefficients (DRCs) in the face of increasing institutional holdings. Their results showed that there were dividend increase of +0.965% and dividend decrease of -1.73% for average excess stock return.

Furthermore, also gave supports that signaling theory plays a significant role in emerging market. 41 cash dividend increase announcements and 39 share dividend announcements from year 1985 to year 1995 were applied as sample in the study. The objective was to test the market response based on those announcements. The outcome showed positive and significant relations and thus the signaling theory was supported by the study.

However, observing this theory provided inconsistent results.argued that dividends increased by the company signal its past but not its future.

2.1.5 Dividend Policy based on Tax Differential Theory

stated that a company will not distribute dividend to its shareholders whenever the dividend is taxed more heavily than the capital gain. Shareholders prefer the firm to retain the profits rather than distribute it to them. These undistributed profits will then be used by the firm to finance its new investments. Therefore, shareholders should invest more in those shares with low or no dividend for future capital gain. He derived the after-tax capital asset pricing model (CAPM), where the before-tax return of stocks is positively related to the tax burden of equity securities in the cross section.

However, examined Brennam’s model and concluded that the tax effect is not relevant. They argued that the capital gain is not affected by the dividend yields either before or after tax. Furthermore, disagreed with Black and Scholes’s result. They criticized their classification of dividend yield. They then applied monthly dividend yield classified into positive yield shares and zero dividend yield shares to prove Brennam model. Their results found that the dividend coefficient was 0.236 and it was highly significant and concluded that Brennam model is relevant.

applied Sharpe CAPM model to examine the relationship between long term dividend yield and share returns. 429 companies in US were used as a sample in his research from the period of 1931 to 1978. He formed six dividend-yield portfolios that consisted of positive dividend-yield and zero dividend-yield companies. His outcomes recommended yield-related tax effect.

2.1.6 Impact of Dividend Policy on Firm Risk

did a research on the determinants of dividend payout ratios. He recommended that growth, agency cost and beta play a role in influencing the optimal dividend payout. He suggested that higher beta coefficients will lead to lower dividend payout. He used the multiple regressions model to test the association of dividend policy with the firm risk with the sample size of 1000 firms from sixty four different industries. It was found that there is significant negative relationship between corporation risk and the dividend payout. He then justified that firms with high beta normally distribute lower dividend because of the higher external borrowing cost.

examined how firms’ size, leverage and dividend influence the risk of ordinary stock. 1000 largest US industrial companies in 1970 were chosen as the sample. Ordinary Least Square regressions model was applied in the study to examine their relationships. They then tested whether there is a significant relationship between alternative risk and firms’ size, leverage and dividend payment by using the correlation matrix. The findings of the study concluded that the firm’s size, leverage and dividend payment have a significant relationship with the corporation risk measures. Besides that, they are listed as the important determinants of firm risk. They showed that the corporations’ risk has a significant negative relation with both the firm size and the dividend yield. However, the corporation risk is negatively affected by the leverage.

examined the relationship between corporation risk and dividend yield based on US stock market. Firms related to 20 years (1960-1979) annual COMPUSTAT files and the monthly share return files from Centre for Research in Security Prices (CRPS) are used as the sample. He applied the simple regression model for the first set of coefficient and used multiple regressions model to obtain the second set of estimates results. Furthermore, he regressed the dividend yield against the beta. The findings stated that there is a significant negative relationship between dividend yield and beta. It concluded that firms with high beta may distribute lower dividend.

2.1.7 Relationship between dividend policy and share price volatility

tested the relationship between dividend policy and share price volatility by applying a different method. Some control variables are added in order to examine the relationship between the dividend policy and the share price volatility. Those control variables consisted of growth, debt, earning volatility and the firm size. These variables clearly affected the share price volatility. They also influence the dividend yield. For example, the earning volatility not only affects the stock price volatility but also influence the firm’s optimal dividend policy. Besides that, assume that the operating risk is unchanged, the debt level may positively affect dividend yield. The firm size may also have an effect on the stock price volatility. Large corporations are mostly more diversified and thus the value of large corporations might be more stable than those of small corporations. Moreover, investors may react irrationally due to the limited public information from small corporations.

Baskin applied the analytical models that link the dividend to the risk of share to his study. The models consisted of information effect, rate of return effect, arbitrage pricing effect and duration effect. The duration effect stated that fluctuation in interest rate has less effect on high dividend yield shares. Corporations that have high dividend yield will have less volatility in stock price since high dividend yield is likely said to be a positive signal of the cash flow. He then applied the Gordon growth model to prove this effect. Furthermore, the rate of return effect by Baskin stated that corporations with low dividend payout and dividend yield can be evaluated more valuable than their assets because of the growth opportunities. By referring to his study, corporations with low dividend payout and dividend yield will have more fluctuation in stock price because the predictions of profit from growth opportunity have more error than the forecast of profit from their assets. He also stated that the excess return is subordinate of price discount rate and the dividend yield since the arbitrage profit will increase by having higher dividend yield. The study also concluded that the share price volatility and the risk of share normally can be controlled by managers when setting the dividend policy. The announcement of dividend distribution may reflect the stability of corporations.

2344 US companies were used as the sample in the study and the time period was from 1967 to 1986. The study concluded that there is a significant negative relationship between the dividend yield and the share price volatility. Besides that, there is also a significant relationship between other variables and the stock price volatility. As a result, Baskin stated that the stock price volatility can be controlled by a corporation’s dividend policy. He also concluded that a decrease in standard deviation of share price trend will lead to an increase in dividend yield. For instance, if the dividend yield rises by 1%, the standard deviation of share price will decrease by 2.5% as stated in the study.

examined the relationship between dividend policy and stock price volatility based on the sample of 73 companies listed in Karachi Stock Exchange (KSE) from the period of year 2003 to year 2008. The panel data was tested by using the fixed effect and random effect models. The result was that there is a significant negative relationship between dividend payout and dividend yield and the stock price volatility. Besides that, the study also concluded that there is non-significant negative relationship between leverage and firm size and the stock price volatility.

examined the relationship between dividend policy and stock price volatility based on the sample of Pakistan companies from five different sectors from year 2005 to year 2009. The data was collected from Karachi Stock Exchange (KSE). They analyze the data by applying multiple regressions model in order to test their associations. The findings stated that there is a significant positive relationship between the dividend policy and the stock price volatility. The result was opposed to Baskin’s findings. The result also stated that the company’s growth affects the stock price volatility negatively.studied the association of stock price volatility with the dividend policy based on UK firms. 123 corporations were chosen as the sample and the time period was from year 1998 to year 2007. They applied the multiple regressions model to examine the association between dividend payout and dividend yield and the stock price. Several control variables were added to their model in the study such as earning volatility, firm size, growth rate and debt level. The findings stated that there is a significant relationship between the dividend payout and the stock price volatility. The result was similar to the findings of study. concluded that there is a negative relationship between the dividend yield and the stock price volatility. As a result, the dividend payout ratio is stated as a major determinant of the stock price volatility according to the study. Besides that, the debt level and the firm size affect the share price volatility the most among the control variables. The findings stated that the firm size affects the share price volatility negatively which was opposed to findings. However, the stock price volatility is affected positively by the level of debt according to the research.

2.1.8 Impact of Dividend Policy on Firm’s Performance

examined the relationship between the dividend policy and the company performance. 406 corporations were used as the sample from the period of year 1989 to year 2000. The results showed that there is a significant relationship between the dividend policy and the company performance. The study applied the model of autocorrelation test in order to examine the relationships. According to the result, changes in dividend strongly connect to current and past performance since positive coefficient of changes in dividend was found after the performance announcement.

Furthermore, also examined the relationship between the dividend policy and the company performance. Twenty five corporations listed on Ghana Stock Exchange (GSE) were chosen as the sample and the time period of the study was from year 1997 to year 2004. Data was collected from those selected firms’ annual reports. In the study, pooled panel crossed-section regression model was applied for the purpose of getting the maximum possible observation. The panel regression model was not similar to the regular cross section or time series regression model since the subscript was doubled to each variable (dividend policy, dividend per share, firm size, debt ratio and growth in sales). The findings showed that there is a significant positive relationship between the dividend policy and return on assets and sales growth. As a result, the research also supports the second school of thought that the company performance is affected significantly by the dividend policy.

Moreover, oncluded that high dividend payment will lead to an increase in market value of stock. Twenty eight corporations from Indian chemical industry were selected as the sample and the time period of the study was from year 1996 to year 2006. By applying stepwise regression model and multiple regressions model in the study, five variables were tested whether they have significant effect on the shareholders’ wealth. These variables consisted of cost of capital, improvement of profit margin, capital structure decision, sales growth and capital investment decisions. The findings stated that there is a significant relationship between the dividend policy and the stock price in those Indian chemical corporations. Therefore, it can be explained that the dividend policy plays an important role to decide the shareholders’ wealth and thus shareholders prefer current dividend to future income.

Chapter 3: Theoretical Framework and Research Methodology¬

3.1 Theoretical Framework

Many past studies have recommended various factors that would affect a company’s share price. In this study, cash dividend will be chosen as the main factor that would affect the firm value. However, there are other factors that could affect the share price as well. Thus, another three factors that were suggested in the past studies most frequently will be selected as control variables in this study. They are net earnings per share (EPS), firm size (total asset) and debt ratio.

The main objective of this study is to examine whether there is a significant relationship between these four factors and company’s share price of selected companies in Malaysia.

Independent Variables

Dependent Variable

Figure 3.1: Theoretical Framework

3.2 Dependent Variable

The dependent variable in the study is the market value of company’s share. Each year’s closing share price will be used as the proxy for this variable.

3.3 Independent Variables

Dividend Policy: The amount of cash dividend (dividend per share) of a company each year

3.4 Control Variables

Three control variables are applied for this research. These variables are chosen as they are recommended in past studies frequently. They are expected to have a great effect on the market value of share and listed as following:

a) Profit: Earning per share (EPS) for each year of a company

b) Firm size: The amount of total asset at the end of each year

c) Debt ratio: The firm’s debt to equity ratio

3.5 Problem statement

The problem statement for this study is:

Is there any significant relationship between dividend policy and the market value of share in selected Malaysian companies?

3.6 Hypothesis Development

• Hypothesis 1:

H0: There is no significant statistical relationship between cash dividend and the market share price of selected companies in Malaysia.

H1: There is a significant statistical relationship between cash dividend and the market share price of selected companies in Malaysia.

• Hypothesis 2:

H0: There is no significant statistical relationship between earnings per share (EPS) and the market share price of selected companies in Malaysia.

H1: There is a significant statistical relationship between earnings per share (EPS) and the market share price of selected companies in Malaysia.

• Hypothesis 3:

H0: There is no significant statistical relationship between firm size and the market share price of selected companies in Malaysia.

H1: There is a significant statistical relationship between firm size and the market share price of selected companies in Malaysia.

• Hypothesis 4:

H0: There is no significant statistical relationship between debt ratio and the market share price of selected companies in Malaysia.

H1: There is a significant statistical relationship between debt ratio and the market share price of selected companies in Malaysia.

3.7 Research Methodology

The main objective for this study is to examine whether there is a significant relationship between the selected four factors and the market value of share of a corporation in Malaysia. The dependent variable for this research is the market value of share of Malaysian corporations. The independent variables for this research are the dividend policy (cash dividend) whereas the control variables are earnings per share (EPS), firm size and debt ratio.

3.7.1 Research Design

In this study, quantitative research is applied which using quantifiable and numeric data. The aim of this study is to identify the association between the independent variables such as cash dividend payment, earnings per share, firm size and debts and the dependent variable (market value of share). The deductive approach is used to formulate hypotheses and measure them empirically. According to quantitative approach uses numerical evidence to form conclusions or to test hypothesis. Besides that stated that quantitative research design is always used to identify the truth value of propositions and enable flexibility in data treatment. For instance, statistical and comparative analysis, data collect and repeatability can define the reliability of the data. Analytical designs are used in this research for the analysis of the relationship between dividend payments and market share price of Malaysian companies to achieve the objectives of this research. Explanatory method is also applied in this research to explain the relationship between the four chosen factors and market value of share of those selected Malaysian companies. Explanatory method is used as it could recognize and meet the purpose of explaining the relationship between the independent variables and the dependent variable.

3.7.2 Research Method

The method of literature search is applied by gathering information from relevant trade publications, newspapers, magazines, annual reports and online data bases.

3.7.3 Data Source

This research uses secondary data. It is collected necessarily as it is used to examine the relationship among the variables and also applied as evidences to support the findings. The secondary data in this study refers to relevant trade publications, newspapers, magazines, corporation annual reports journals, relevant articles, Google Scholar and online data bases (EBSCO). The sources enable the collection of the market price of share of those selected companies. The published accounts of those chosen companies can be obtained from Bursa Malaysia website. Furthermore, Statistical Package for Social Sciences (SPSS) program is used to analyze and test the hypotheses by calculating the means, standard deviations, correlation coefficient.

3.7.4 Sampling Measuring Scale

The scale applied to examine the relationship among those variables is interval scale. This is because it is assumed to have equidistant points between each of the scale elements. Interval scales are also scales which are defined by metrics such as logarithms. Examples of parametric statistical techniques used in this study are Pearson correlation-r and significant (2-tailed).

3.7.5 Data Collection

Several stages are carried out in the data collection process. The first stage is to identify those selected corporations for which data are available on the announcement of the dividend date. The number of selected corporations is 103. The second stage is to collect relevant data for those selected corporations. The annual data (collected through Bursa Malaysia website) of the companies listed at Bursa Malaysia for the period of 2008 to 2012 is identified.

A total of 103 listed Malaysian companies that declared the dividend payment from the year 2008 to the year 2012 are selected as the sample for this research. A period of five years (2008 to 2012) is applied in this research. The dividend payment refers to cash dividend. The requirement is that those selected companies should have at least paid cash dividend once over the past five financial years.

3.8 Research Model

Multiple regression model and Pearson correlation analysis are used in this study in order to analyse the relationship between independent variables and dependent variable. (Baskin, 1989)

3.8.1 Multiple Regression Model

This model which primarily connects the stock price to dividend payment has been developed by other variables In this study, these variables consist of earnings per share, firm size and debt ratio.

The dependent variable (share price) is regressed on the main independent variable (cash dividend). Therefore, the equation of the multiple regressions model is given as below:

Pt = b0 + b1Dt + et

Where,

Pt = Share price for firm t

Dt = Dividend per share for firm t

et = Error term

Furthermore, earnings per share, firm size and debt ratio may also affect the dividend policy and the share price The stock price of small firms is usually more volatile than large firms since small firms may have less diversification and information available to investors. Moreover, leverage could have effects on the stock price due to operating risks. Also, the existence of asymmetric information links the borrowing policy to the dividend policy. As a result, the equation of the regression model is expanded as below:

Pt = b0 + b1Dt + b2EPSt + b3SIZEt + b4DEBTt + et

Where,

Pt = Share price for firm t

Dt = Dividend per share for firm t

EPSt = Earnings per share for firm t

SIZEt = Total asset for firm t

DEBTt = Debt Ratio for firm t

et = Error term

3.8.2 Pearson Correlation Analysis

Pearson Correlation Analysis is applied in this study for the analysis of relationship between independent variables and the dependent variable. This model explains the direction and relative strength of a linear relationship between two numerical variable. The values of the coefficient of correlation range from -1 for a perfect negative correlation to +1 for a perfect positive correlation.

In this study, the two variables are dividend per share distributed by those selected companies during the study period and their share prices. The calculation of the coefficient of correlation is done for each of 150 selected firms and the average calculated from the individual result. The formula is presented as below:

Where

rxy = Correlation of x and y

n = Number of items or measurement

x = First measurement

y = Second measurement

3.9 Limitation of the research

Several limitations were occurred in the process of completing this research. Those limitations may influence the accuracy of the result of findings. The major limits of this research are the size of the sample and the period of time of the study used. Most research strives for the largest possible proportion of its targeted community and for the longest period of time possible in order to get accurate and reliable results that can be generalized confidently. Excessive volatilities were occurred in the Bursa Malaysia during the year 2007 to year 2008 in term of data volatilities. The Malaysian share market was affected by the US subprime crisis and led to a rapid sell-off during the period. It will lead to a decrease in the share price due to excessive supply. Besides that, the stock market was also affected by the instability of political issues during that period. The 12th National General Election caused the FTSE Bursa Malaysia to decrease by nearly 130 points in year 2008. As a result, it is believed that the findings of this research may be affected by these stated volatilities due to the increase of variances in stock prices. Moreover, the inadequacy of research skills, expertise and limited research tools used may also influence the reliability of the findings.

Chapter 4.0 Data Analysis

4.1 Regression Assumption Tests

The regression assumption tests are applied in this study for the reliability of the model before analyzing the data through regression model. Requirements of normality, linearity and collinearity must be met before applying the regression model to analyze the data.

4.1.1 Collinearity Test

Collinearity requires that the independent variables should not correlate with each other. The existence of collinearity causes independent variables cannot be completely independent. This will occur when one independent variable strongly correlates with another and usually the regression model is not added with new information. Besides that, the model outcomes will be violated by the correlation due to the problem of isolating the effect of strong correlation between the selected independent variables.

One of the ways applied to measure the collinearity is to define the variance inflationary factor (VIF) for every independent variable. The formula of VIF is as below:

VIF = 1/ (1-R2)

The decision rule of this measurement is that the collinearity has no impact on the association between the independent variable and the dependent variable if the independent variable’s VIF is equal to one. Furthermore, an independent variable is said to have a collinearity with other independent variables if the independent variable’s VIF coefficient is larger than five.

The collinearity test between the independent variables (DPS, EPS, total asset and debt ratio) is applied in this study in order to define whether there is a collineary problem for each independent variable with a significant impact on the association between the independent variables and the dependent variable by using the SPSS software.

Table 4.1 Variance Inflationary Factor (VIF)

The table 4.1 shows the VIF coefficient between the independent variables (DPS, EPS, total asset and debt ratio) based on the sample of 103 Malaysian companies from 2008 to 2012. The DPS has a VIF coefficient of 3.265 and the VIF coefficient of EPS is 4.13. The TA and DEBT have VIF coefficient of 1.586 and 1.088 respectively. According to the results, none of independent variable’s VIF coefficient is equal or greater than five. Therefore, it concludes that no collinearity is associated between the independent variables and the requirement of collinearity for regression model is met.

4.1.2 Normality Test

Another important assumption of the regression model is the normality assumption. The assumption of normality requires that all variables must be normal distributed. Regression analysis is fairly robust against departures from the normality assumption. The regression equation will not be affected seriously as long as the distribution of the variables is not abnormally different from the normal distribution. In other words, the outcomes will be more accurate if the variables follow the normal distribution.

In this study, the normality of the independent variables (DPS, EPS, total asset and debt ratio) and the dependent variable (share price) will be tested through observing the skewness and kurtosis and Shapiro-Wilk test. The skewness statistic assesses the area to which a data set is not normal distributed. The value of zero of the skewness means the data set is normally distributed. The distribution is right skewed if the skewness has a positive value whereas the distribution is left skewed if the skewness has a negative value. The kurtosis statistic assesses the relative concentration of values in the middle of the distribution of a data set as compared to the tails. The value of zero of the kurtosis indicates the data set is normally distributed. The distribution is flatter than the bell-shaped distribution if the kurtosis has a negative value whereas the distribution is sharper than the normal distribution if the kurtosis has a positive value.

Descriptives

 

Statistic

Std. Error

SP

Skewness

.161

.238

 

Kurtosis

.279

.472

DPS

Skewness

.164

.238

 

Kurtosis

-.211

.472

EPS

Skewness

-.491

.238

 

Kurtosis

.901

.472

TA

Skewness

-.287

.238

 

Kurtosis

-.641

.472

DEBT

Skewness

-.048

.238

 

Kurtosis

.721

.472

Table 4.2 Skewness and Kurtosis for all variables (SP, DPS, EPS, total asset and debt ratio)

Table 4.2 shows the value of skewness and kurtosis for the independent variables (DPS, EPS, total asset and debt ratio) and the dependent variable (share price) based on 103 selected Malaysian companies from 2008 to 2012. All the value of skewness and kurtosis are computed by SPSS software. As shown in the table, the variable of share price (SP) has a skewness of 0.161 and a kurtosis of 0.279. The DPS has a skewness of 0.164 and a kurtosis of -0.211. The EPS has a skewness of -0.491 and a kurtosis of 0.901. The TA has a skewness of -0.287 and a kurtosis of -0.641. The DEBT has a skewness of -0.048 and a kurtosis of 0.721. As a result, it can conclude that all the variables are estimated to be close to normal distribution. Therefore, this allows the possibility of drawing results of the regression model.

The other method to test the normality of variables is to run Kolmogorov-Smirnov and Shapiro-Wilk test. Kolmogorov-Smirnov test is used if there are more than 2000 variables. Therefore, Shapiro-Wilk test is appropriate to apply in this study to test the normality.

Tests of Normality

 

Kolmogorov-Smirnova

Shapiro-Wilk

Statistic

df

Sig.

Statistic

df

Sig.

SP

.066

103

.200*

.991

103

.709

DPS

.047

103

.200*

.995

103

.962

EPS

.072

103

.200*

.976

103

.059

TA

.087

103

.051

.976

103

.054

DEBT

.099

103

.015

.978

103

.077

*. This is a lower bound of the true significance.

a. Lilliefors Significance Correction

Table 4.3 Kolmogorov-Smirnov and Shapiro-Wilk test

Table 4.3 displays result of the Kolmogorov-Smirnov and Shapiro-Wilk test for each variable. Assume the null hypothesis (H0) states that the variable is normally distributed and the alternative hypothesis (H1) states that the variable is not normally distributed. As shown in the table, SP has a p-value of 0,709 and DPS has a p-value of 0.962. EPS has a p-value of 0.059 and TA has a p-value of 0.054. Lastly, DEBT has a p-value of 0.077. At the 0.05 level of significance, the p-value of all the variables are greater than 0.05. Therefore, we fail to reject the null hypothesis (H0) and it can conclude that all the variables are normally distributed.

4.1.3 Linearity Test

Furthermore, another assumption for regression model is linearity. It indicates that the relationship between the independent variables and the dependent variable must be linear. In this study, a residual plot is used in order to test the linearity between the independent variable and the dependent variable. The regression residuals versus the independent variable values are marked in the residual plot. Generally, the residual plot is applied to measure whether the regression line is fit. There is no apparent cluster of positive residuals or a cluster of negative residual if linearity occurs.

By using SPSS program, the residual plots for each independent variable are presented as below:

Figure 4.1 Residuals versus DPS

Figure 4.2 Residuals versus EPS

Figure 4.3 Residuals versus TA (total asset)

Figure 4.4 Residuals versus DEBT

By observing the above figures, all the variables meet the assumption of linearity. Therefore, it can conclude that the regression model is suitable for the data of this study according to all the previous assumption tests. On the other hand, it can also conclude that the multiple regressions model can be counted on to explain the dependent variable.

4.2 Descriptive Analysis

Descriptive Statistics

 

N

Minimum

Maximum

Mean

Std. Deviation

SP

103

.41

49.40

5.5522

6.61929

DPS

103

.01

2.61

.2557

.34634

EPS

103

.02

2.66

.3926

.40406

TA

103

.27

10.98

3.2583

2.20385

DEBT

103

0.00

14.04

1.1098

2.37569

 

 

 

 

 

 

Table 4.4 Descriptive Statistic

Table 4.4 presents the descriptive statistic of all the variables which are applied in this study based on 103 selected Malaysian companies from the period of 2008 to 2012. It shows the minimum, maximum, mean and standard deviation. As shown in the table, the minimum of the share price (SP) is 0.41 and the maximum is 49.40. It has a mean of 5.5522 and the standard deviation is 6.61929. However, the outcome found by Salih and Alaa (2010) is different from this study. According to their study on how dividend policy affects share price, it has a greater mean of 400.0213 and a higher standard deviation of 971.2053 based on the sample of 362 UK companies. Besides that, the study of Zuriawati (2012) has a smaller mean of 0.9441 and a standard deviation of 0.4047 based on the sample of 106 Malaysian construction and materials firms.

The minimum of dividend per share (DPS) is 0.01 and the maximum is 2.61. It has a mean of 0.2557 and a standard deviation of 0.34634. It indicates that the share price will increase by 0.08856% (0.2557 x 0.34634) with one percent of the standard deviation of DPS increases. The result is slightly close to Mohammad and Ardekani (2012)‘s findings. For the dividend, their result has a mean of 0.03805 and a standard deviation of 0.03618 based on the sample of 84 consumer product manufacturing firms. It means an increase by one percent of the standard deviation of the dividend will lead to an increase in share price by 0.001376%. Moreover, Salih and Alaa (2010) found a higher mean of 5.057854 and standard deviation of 357.9381 for the dividend variable based on 362 UK firms.

Next, the table shows that the minimum of earnings per share (EPS) is 0.02 and the maximum is 2.66. It has the mean of 0.3926 and a standard deviation of 0.40406. It indicates that the share price will increase by 0.15863% (0.3926 x 0.40406) when the standard deviation of EPS increases by one percent. According to the study by Salih and Alaa (2010), its EPS variable has a higher mean of 0.119084 and a standard deviation of 1.443107. It means the share price will increase by 0.17185% with one percent increase in standard deviation of the EPS variable based on the sample of 362 UK companies.

Furthermore, the total asset (TA) is measured by using the natural logarithm of total asset. The minimum of total asset (TA) is 0.27 and the maximum is 10.98. It has a mean of 3.2583 and a standard deviation of 2.20385. It means the share price will increase by 7.18% with one percent increase in the standard deviation of TA. However, the study of Salih and Alaa (2010) found a higher mean of 5.50564 but a lower standard deviation of 0.97822 based on the sample of 362 UK companies. It indicates that one percent increase in standard deviation of the total asset variable will lead to an increase in share price by 5.3857%.

Lastly, the debt ratio (DEBT) is measured by total debt over total equity. The minimum of DEBT is 0 and the maximum is 14.04. It has a mean of 1.1098 and a standard deviation of 2.37569. It means the share price will increase by 2.6365% (1.1098 x 2.37569) with one percent increase in standard deviation of DEBT. However, Mohammad and Ardekani (2012) found a lower mean of 0.09347 and a standard deviation of 0.5672 for debt ratio variable. Besides that, the study by Zuriawati (2012) also has a smaller mean of 0.6352 and a lower standard deviation of 0.8624 for debt ratio variable.

According to the results in table 4.4, the total asset (TA) and the debt ratio (DEBT) have a bigger effect on share price whereas the DPS and EPS have a small effect on share price.

4.3 Regression Analysis

4.3.1 Model Ability (R, R-Squares and F-test)

Model Summary

Model

R

R Square

Adjusted R Square

Std. Error of the Estimate

1

.888a

.789

.781

.1795925

a. Predictors: (Constant), DEBT, EPS, TA, DPS

Table 4.5 Coefficient

As shown in table 4.5, the R refers to correlation coefficient. It is applied to assess the degree of correlation between the independent variables and the dependent variable. Generally, the value of range is between -1 and 1. It indicates a strong positive correlation between the independent variable and the dependent variable if the value is 1. However, it indicates a strong negative correlation between the independent variable and the dependent variable if the value is -1. In this study, the independent variables are dividend per share, earnings per share, total asset and debt ratio and the dependent variable is share price. The table shows that the R is 0.888. The result is close to 1 and thus it means there is a strong relationship between the independent variable (DPS, EPS, TA and DEBT) and the dependent variable (SP).

Furthermore, by assessing the value of R square, the dependent variable can be explained by the proportion of the independent variables. R square refers to coefficient of determination. It is used to measure the proportion of variation of the dependent variable that is explained by the independent variable in the regression model. As shown in table 4.5, the R square is 0.789 (78.9%). It means 78.9% of the variation in dependent variable is explained by the variation in the independent variables. In other word, the DPS, EPS, total asset and debt ratio in the model explain 78.9% of the share price. However, the model cannot explain the remaining 21.1%.

ANOVAa

Model

Sum of Squares

df

Mean Square

F

Sig.

1

Regression

11.843

4

2.961

91.798

.000b

Residual

3.161

98

.032

Total

15.004

102

a. Dependent Variable: SP

b. Predictors: (Constant), DEBT, EPS, TA, DPS

Table 4.6 ANOVA

Moreover, the overall F test is applied in order to test whether there is a significant relationship between the entire set of independent variables and the dependent variable by comparing the calculated F as shown in table 4.6 with its critical value. If the calculated F is larger than the critical value, it indicates that the R square is statistically significant and therefore the result from the model can be used. The calculated F as shown in table 4.6 is 91.798 which is greater than its critical value (2.38). As a result, it can conclude that there is a statistically significant relationship between the independent variables and the dependent variable at the 0.05 level of significance.

4.3.2 Multiple Linear Regression

Coefficientsa

Model

Unstandardized Coefficients

Standardized Coefficients

t

Sig.

B

Std. Error

Beta

1

(Constant)

1.028

.062

16.516

.000

DPS

.295

.061

.347

4.840

.000

EPS

.485

.074

.519

6.586

.000

TA

.184

.071

.148

2.609

.011

DEBT

.022

.034

.032

.661

.510

a. Dependent Variable: SP

Table 4.7 Regression

Table 4.7 displays the multiple regression model result between the independent variables (DPS, EPS, TA and DEBT) and the dependent variable (SP) based on the sample of 103 selected Malaysian companies from 2008 to 2012. Those hypotheses of this study are tested in the next section:

4.3.2.1 Dividend per share (DPS)

Hypothesis 1:

H0: There is no significant statistical relationship between cash dividend and the market share price of selected companies in Malaysia.

H1: There is a significant statistical relationship between cash dividend and the market share price of selected companies in Malaysia.

According to table 4.7, the p-value for DPS is 0.000. At the 0.05 level of significance, the null hypothesis (H0) can be rejected and the alternative (H1) is accepted because the p-value = 0.00 < 0.05. Therefore, it can conclude that there is a significant statistical relationship between the cash dividend (DPS) and the market share price of selected companies in Malaysia at 95% confidence level.

4.3.2.2 Earnings per share (EPS)

Hypothesis 2:

H0: There is no significant statistical relationship between earnings per share (EPS) and the market share price of selected companies in Malaysia.

H1: There is a significant statistical relationship between earnings per share (EPS) and the market share price of selected companies in Malaysia.

According to table 4.7, the p-value for EPS is 0.00. At the 0.05 level of significance, the null hypothesis (H0) can be rejected and the alternative (H1) is accepted because the p-value = 0.00 < 0.05. Therefore, it can conclude that there is a significant statistical relationship between the earnings per share (EPS) and the market share price of selected companies in Malaysia at 95% confidence level.

4.3.2.3 Total Asset (TA)

Hypothesis 3:

H0: There is no significant statistical relationship between firm size and the market share price of selected companies in Malaysia.

H1: There is a significant statistical relationship between firm size and the market share price of selected companies in Malaysia.

According to table 4.7, the p-value for TA is 0.011. At the 0.05 level of significance, the null hypothesis (H0) can be rejected and the alternative (H1) is accepted because the p-value = 0.011 < 0.05. Therefore, it can conclude that there is a significant statistical relationship between the firm size (TA) and the market share price of selected companies in Malaysia at 95% confidence level.

4.3.2.4 Debt ratio (DEBT)

Hypothesis 4:

H0: There is no significant statistical relationship between debt ratio and the market share price of selected companies in Malaysia.

H1: There is a significant statistical relationship between debt ratio and the market share price of selected companies in Malaysia.

According to table 4.7, the p-value for DEBT is 0.510. At the 0.05 level of significance, the null hypothesis (H0) is accepted and the alternative (H1) can be rejected because the p-value = 0.510 > 0.05. Therefore, it can conclude that there is no significant statistical relationship between the debt ratio and the market share price of selected companies in Malaysia at 95% confidence level.

4.4 Pearson Product-Moment Correlation Analysis

Correlations

 

SP

DPS

EPS

TA

DEBT

SP

Pearson Correlation

1

.775**

.853**

.534**

-.035

Sig. (2-tailed)

 

.000

.000

.000

.726

N

103

103

103

103

103

DPS

Pearson Correlation

.775**

1

.749**

.312**

-.180

Sig. (2-tailed)

.000

 

.000

.001

.069

N

103

103

103

103

103

EPS

Pearson Correlation

.853**

.749**

1

.523**

-.065

Sig. (2-tailed)

.000

.000

 

.000

.513

N

103

103

103

103

103

TA

Pearson Correlation

.534**

.312**

.523**

1

.195*

Sig. (2-tailed)

.000

.001

.000

 

.049

N

103

103

103

103

103

DEBT

Pearson Correlation

-.035

-.180

-.065

.195*

1

Sig. (2-tailed)

.726

.069

.513

.049

 

N