Determinants of Dividend Pay-Out Policy, Essay Example

Background

Dividends are defined by Investopedia as “a distribution of a portion of a company’s earnings, decided by the board of directors, to a class of its shareholders”. The dividend is most often quoted in terms of the fixed amount each share receives (dividends per share – DPS). Investors seek to see returns and growth in companies that they invest in, in the form of dividends and rise in share price respectively. Dividend policy informs managers’ payout recommendations regarding the way that the board of directors should distribute company profits back to the shareholders as dividend and/or be retained as reinvestment in the company to fund future growth.

There are several theories that guide dividend policy. In terms of the residual theory of dividends; companies give precedence to its financing needs, and only pay dividends out of remaining earnings which the company has no immediate use of. The basic assumption of residual dividend theory is that shareholders want the company to retain earnings if reinvesting them can generate higher rates of return than the shareholders could obtain by reinvesting their dividends. Gitman (2009) asserts further that dividend payout decisions must not be made in isolation, but only be made after investment and capital structuring decisions. Therefore, according to the residual theory, leverage is a factor that must be considered to inform dividend payout policy.

According to Boundless Finance (2014), Miller and Modigliani’s dividend irrelevance theorem postulate that a company’s dividend policy is not relevant, because shareholders are ultimately indifferent between receiving returns from dividends or capital gain. (Malik et al 2013:35) stated that “the firms’ value and shareholders’ wealth are not related to the decision whether or not the firm pays dividend”. This theory implies that companies attract like-minded investors, therefore creating a clientele effect. Kawano (2014:1) explains this effect as a process when “investors naturally sort into equity holding classes based on their dividend payout ratios”.  This theory is not fully applicable to South Africa, because interest on debt can be claimed as a tax deductible expense, thereby increasing the company value. In addition, according to this dividend irrelevance theory; debt or leverage is a non-factor in dividend policy, which is in contradiction with the residual theory of dividends.

In their study of factors influencing dividend payout decisions of firms listed on Karachi Stock Exchange, Malik et al (2013) managed to establish that leverage, earnings per share, liquidity, and size are related to dividend payouts. This finding  also contradicts the Miller and Modigliani’s theorem, and supports the dividend irrelevance theory. The above mentioned study found that company size is related to dividend payout, thereby reinforcing the dividend relevance theory. The dividend relevance theory states that there is a direct link between dividend policy and market value, supporting “the bird in hand argument” that investors see current dividends as less risky than future dividends or capital gains (Gitman 2009).

Another approach, the signal theory states that companies that make public announcements of changes in dividend pay-outs as signals to the market, thereby using the statements as indicators of the companies’ future forecasts. For the South African mining market, this trend is clearly visible. An article by Ferreira-Marques (2013 para. 1) mentions the example when “Glencore Xstrata wooed investors on its debut as a combined miner and trader on Friday, with promises of dividend payments and signals of aggressive cost cuts that meant it could beat a planned synergy target of $500m”. Thus, found to be the most relevant to the South African mining industry, this theory is found to be the most applicable to use in this study.

Lastly, Denis and Osobov (2008) created a study of international determinants of dividend policy. Their findings cast doubt on signalling and clientèle effects of dividend theories. They confirmed an alternative theory, called the life cycle theory that states “in their early years, firms pay few dividends because investment exceeds internally generated capital. In later years internal funds exceed investment opportunities so firms pay out the excess funds to mitigate misuse of free cash flows” Denis and Osobov (2008:62). The mining industry is very capital intensive, and this theory could help to explain the varying dividend payout practices used by different companies.

The current literature regarding dividend policy is very vast, and covers most countries, stock exchanges and various industries, especially in the Western and Middle Eastern economies. As Porta, Lopez-de-Silances, Schleiffer, and Vinshy (2000) confirm, dividend policies depend on multiple variables, and most importantly on the country of operation.

Macroeconomics, socio-economic, and other environmental factors are unique in  South Africa, and many theories of policy are not applicable to the mining industry. It is clear from the above mentioned studies that countries and companies with high mining exposure have also not been thoroughly explored in previous studies.

The very few studies (conducted in the developing East and Asian countries of Malaysia, Pakistan and Indonesia are quite relevant to the South African economy, as it is also a developing country. Therefore, the below study aims to fill the gap in research and make a contribution to the body of knowledge in the South African context, particularly in reference to the mining industry, which is still a major contributor to the gross domestic product (GDP) of the country.

There are factors that impact on the ability of the company to make dividend payout decisions. Determinants of dividends have been established to be amongst others company size, growth opportunities and profitability by Denis and Osbov (2008). Further, examining the South African gold mining industry, Ringane and Makoni (2014) found that the size of the company has a great influence on dividend payment strategies. Foreign ownership was also a factor that determined the company’s dividend policies, while managerial ownership was in weak negative correlation with the existence and size of dividend payments.

The factors were further categorised by Malik et al (2013), and the author added  internal factors, as  investment opportunities, profitability, and liquidity, as well as external factors; comprising company growth, stability, technology, and macroeconomics to the list of conditions influencing dividend policies.

As a young developing democracy, South African companies have to compete with developed and developing countries’ mining related investments in the country. It is therefore very important and imperative that dividend decisions are made based on research of conditions. The below study is carried out to support companies in establishing dividend strategies. Through determining country specific factors in the mining industry, the researchers can provide a framework for the South African mining industry in order to make the country more attractive for investors in the future.

Problem statement

The purpose of this study is to investigate the factors that contribute to the JSE – listed companies in the mining industry decisions regarding dividend payout for the period between 2004 and 2013. The secondary objective is to establish if there is any correlation to between country-specific, company-specific, and industry-specific factors and dividend payout.

The data used for the current research covering dividend decisions of 53 mining companies was extracted from the McGregor Financial Analysis System (FAS) financial database. The main purpose of the research is to determine whether there is a relationship between internal financing factors and dividend payout decisions. Specifically the effects of growth, profitability, leverage, liquidity and company size are investigated, to determine their impact on dividend payout decisions.

For this study, the dividend payout is used as the dependent variable with the earnings per share, debt to equity ratio, net profit margin, current ratio and company size being independent variable used in a statistical analysis.

Objective of the research

The objectives of this study were:

• To determine the influence of earnings per share (growth rate) on dividend payout among JSE-listed mining companies.
• To determine the influence of net profit margin on dividend payout (profitability) among JSE-listed mining companies.
• To determine the influence of debt to equity ratio (leverage) on dividend payout among JSE-listed mining companies.
• To determine the influence of current ratio (liquidity) on dividend payout among JSE-listed mining companies.
• Lastly, to determine if there is a correlational relationship between earnings per share, net profit margin, debt to equity ratio and current ratio and dividend payout among JSE-listed mining companies.

The report is structured as follows: Chapter 1: Introduction giving background to the study, Chapter 2: Literature review and theoretical framework of the determinants of dividend policy, Chapter 3: Research Methodology describing methods utilised to collect data, Chapter 4: Research Findings describing results and findings, Chapter 5: Conclusions of the study and recommendations thereof.

Theoretical framework and literature review

There has been a lot of research that has gone into answering the question why dividends are paid out. Malik et al (2013:42) in their study concluded that “dividend paid at the appropriate time has a positive impact on a reputation of a company”. The findings of the authors’ research include that there is a negative correlation between firm growth and dividend payout. The study concluded in Pakistan, a developing country showed that liquidity, profitability, earnings per share, and company size increased the probability of the firm paying dividends.

In a study by Denis and Osobov (2008) the authors discovered that bigger, profitable firms with higher retained earnings were the ones most likely to pay dividends. This study supported lifecycle theories, but cast doubt on signalling and clientèle theories. Denis and Osobov’s (2008) findings were further corroborated by Meyers and Bacon (2004) when the authors found that firms with larger profits and cashflows are more likely to pay dividends. In a previous research, however, using the residual theory, Kania and Bacon (2005: 105) found that “sales growth and expansion related negatively to the dividend payout ratio”.

A study by Baker and Powell (2009) have indicated that most Indonesian executives considered the stability of earnings and the expected future earnings to be the most important determinants of dividend policy with the policy also affecting firm value. This study supports signalling and life cycle theory for payment of dividends. Meyers and Bacon (2004) also established that dividends cashflow had a signalling effect. Their study indicated that earnings, access to capital, and high growth had a direct impact on dividend payout.

A recent study of Pakistani financial and non-financial firms by Malik et al (2013) showed that liquidity, leverage, EPS, and size are positively related to dividend payouts, whereas growth opportunities decrease the possibility of a payout. A similar study by Ardestani et al (2013), agreeing with the Dividend Irrelevance Theory, also found that investment opportunity and financing are important factors in determining dividend policy in Malaysian industrial firms, with profitability and risk influencing those decisions.

A paper by Allen and Michaely, (2002) focusing on the dividend payout policies survey of American companies, identified taxes, asymmetric information, and dividend transaction cost as some of the factors that affect dividend payout decisions. The findings of the study are consistent with Miller and Modigliani’s debt irrelevance theory, assuming that payout policies are not motivated by “the firms’ to signal their true market value, but by management in their effort to reduce over-investment in the company. The study highlighted that there is a current change of trend in dividend policy among most American companies. American firms now decide to pay shareholders in the form of share repurchases than dividend payout, because of the high transactional and taxation costs associated with dividends. Even though this study is based on American companies, is relevant to the current research, because most mining companies operating in South Africa are dual-listed, and are influenced by global trends in the market. The limitation of the study is that it surveyed only US companies, and tracked changes over the last 5 decades, instead of focusing on recent trends. Some of the practices examined have been discontinued, and there are a lot legal and accounting practices and laws that have been introduced in most markets due to structural changes in recent years.

The Resource Investing News’ article by Chad Fraser (2012), the author concludes that most investors he interviewed are attracted to mining stocks/shares because they pay out dividends despite the industry’s high volatility. He suggests that dividend payout helps lowering the risk of investing in mining and metals company, because it gives investors some degree of security. In this case, dividend payout plays a signalling role indicating that the company will have sufficient reserves and earnings in the future and can afford to pay out dividends. This finding supports the signaling theory’s applicability in the mining industry. Fraser (2012) further concludes that a dividend paying mining company will avoid reducing or not paying out any dividend because it might send negative signals about the financial wellbeing of the company. The limitation of Fraser’s (2012) study is that it involved very few companies and the sample size restricts the findings to be applied to the general global mining population. The study was focused on Toronto and New York Stock Exchanges listed mining companies, but the trends highlighted can be further investigated in South African mining companies and thereby making this study very relevant to the current research topic.

Firer et al (2008) study of dividend policy in South Africa concluded that companies’ policies in the country are impacted by the following factors: EPS, free float and share liquidity, credit ratings and capital structure, taxation and takeover resistance. Of significance is that secondary tax on companies (STC) was only introduced in 1993, and it had a strong influence on dividend payout. This impact of secondary tax has not been studied intensively, and examining the relationships between taxation and dividend policies can be incorporated in the current research, in order to establish if the mining industry has been significantly influenced by the introduction of STC. The share repurchase option was also introduced to the South African economy in 1999, as a complementary alternative form of payout of dividends to the shareholders. Firer et al (2008) further established that South African managers’ behaviour was similar to that of their United States counterparts towards dividend/share repurchase policy. Both manager attitudes are considered to be conservative. Managers of both South African and American companies set dividends payout targets so that they can have breathing space and avoid payout cuts in the future. signaling theory is therefore hard at play when SA managers make their payout decisions. This similarity is significant, as relevant US studies can be used to analyse management decisions related to dividend payments among SA listed companies, and therefore, these studies can be used for research purposes. For this study, the researchers highlighted one of the main possible limitations of research: the low response rate that they experienced.  This is why structured interviews should complement any further research. This will definitely be incorporated in my research methodology so that more reliable results can be delivered.

When Marks (2001) surveyed directors of SA listed companies, most respondents emphasized the importance of communicating dividend decisions to shareholders. This further supports the application of signaling theory by most managers in the South African markets. Neill, Hamman and Smit (2001) also came to the same conclusions when they studied SA listed companies. According to their research, managers preferred to communicate with shareholders, and they had a conservative approach towards setting dividend payout ratios, thereby avoiding the need to communicate future reductions or discontinuation of dividends. These studies are both relevant and applicable to the current research, because they focus on SA listed companies.

From the above literature review, it is apparent that the majority of studies somewhat indicate the validity of the signalling, clientèle, and life cycle dividend theories, though some elements of the findings might be contradictory.

Other factors affecting dividends payout were further researched by Porta et al (2000) in over 33 countries including South Africa. His study found that minority shareholders often flexed their muscles on company managers so that they could extract dividends as per their legal rights as investors. The authors observed that countries offering legal protection tend to pay out higher dividends. This study will be very helpful when looking at the developing South African legal and commercial landscape. This study did not confirm signaling theory, nor did it find that tax had an effect on dividend payout.

Research hypotheses

The Null hypotheses for the study are as follows:

1. There will be no difference in company size and dividend payout as measured by mean scores of market capitalisation and dividend per shares issued respectively of JSE-listed mining companies.
2. There will be no difference in company growth and dividend payout financial performance, as measured by mean scores for capital expenditure and dividend per share issued of JSE-listed mining companies.
3. There will be no difference in environmental factors and dividend payout as measured by mean scores for factors and dividend per share issued of JSE-listed mining companies.
4. There is no relationship between company growth, environmental factors and dividend payout of small, medium and large capitalized JSE-listed mining companies as measured through correlation coefficients of the mean scores.

The corresponding Research hypotheses for the study:

1. Company size as measured by mean scores of market capitalisation is different.
2. Dividend payout as measured by mean scores of dividend per shares issued of JSE-listed mining companies is different.
3. Environmental factors as measured by mean scores of the factors exist and are different for mining companies.
4. There is a relationship between size, growth, environmental factors and dividend payout of JSE-listed mining companies.

Research Methodology

The aim of this research is to establish the factors that contribute to JSE-listed mining companies’ decisions to pay out dividends, with particularly internal financing factors and how external environmental factors affects those decisions. In this study, the authors would like to establish if and how factors like company size, company growth, and financial performance impact on dividend payout decisions. The research is historical because it is based on past company financial performance.

A quantitative methodology was followed, because this method can be helpful in establishing or inferring existence of statistical relationships among the factors (Creswell 2003) when reviewing historical financial data. According to Salkind (2012), the benefit of mixed approach is that results from all sources can be examined separately and together to confirm the findings of the research. Therefore, in addition to using a numerical database, a close ended survey was also done to confirm the quantitative study.

Salkind (2012)  describes quantitative correlational research as one that examines relationship between quantitative variables; which gives indication how two or more variables are related to one another or how well a specific outcome can be predicted by the other variable. A recent study of dividend payout of Pakistani financial and non-financial firms by Malik et al (2013) also employed correlational research methodologies. Thus, the authors of the current study also chose quantitative correlational approach, because it is effective in establishing if there is a relation between above mentioned factors and dividend payouts. When employing the quantitative approach, the study will use quantitative financial performance data collated from annual company reports and correlate it with dividend payout of the respective mining firms to establish existence of a relationship.

In a study of dividend policy in Indonesia done by Baker and Powell (2009)  survey/questionnaire methods were used, and the authors surveyed executives to learn their views about factors that influenced dividend policy. This method is well established and was therefore also used in the current study with the aim of understanding how they would have considered these factors in making dividend payout decisions.

Sampling

The target population for this study is all 53 mining companies listed on Johannesburg Stock Exchange (JSE) for the period between 2002 and 2012. The JSE has a total of almost 400 companies listed on the main board (large capitalisation), and the Alternative Exchange (AltX) board featuring small to medium sized capitalised companies. Our study focus is on 45 large capitalised mining companies listed on the main board, and 8 small and medium sized high growth mining companies listed on the AltX. I-Net Bridge database, which provides real time and historical financial performance data was used as our main source of information. This database will be cross checked with company reports.

The stratified random probability sampling method was employed to generate a sample. The main advantage of stratified sampling according to Kotze (2007) is that it captures key population characteristics in the sample. This method was chosen to ensure that the sample is chosen proportionally from both the main board and AltX stratified according to company size to limit bias and ensure the correct representation of the main population. Utmost care was taken to ensure that the sample is representative of the population so that any observations made for the sample should also hold true for the larger population, and researchers will be able to make generalisations to different populations of similar characters (Salkind 2012). Most recently the stratified random sampling was also used by imminent finance researchers Ardenstani et al (2013) when they studied dividend payout policies of industrial companies in Malaysia.

Due to time and financial constraints, it is not always possible or practical to study the entire population. In most cases, a sample of 10% of the population is sufficient. However, in this case, with our population of 53 companies the research would be restricted to 6 companies, and would be insufficient for generalisation purposes. The number of participating companies selected was hence 30, because it is a generally accepted size required to ensure minimum statistical error (Salkind 2012). Therefore, a computer generated random list of the proportionally stratified 30 main board and AltX mining companies was sourced from the I-Net Bridge database. As some respondents have not submitted their answer yet, the response rate will be confirmed in the final research report.

Data collection

In this study, data will be collected threefold from I-Net Bridge, JSE, and company annual reports. I-Net Bridge Financial Analysis System (FAS) is a database that has more than 20 years data on financial performance of JSE listed companies. This method of collecting data from financial databases was most recently used by Arderstani et al (2013) in their study of dividend policy of Malaysian industrial companies, and has also been previously used by Myers and Bacon (2004) in their study of determinants of dividends for companies listed on the New York Stock Exchange.

The JSE and company reports will then be used to corroborate information and to source additional company information if required. Mining companies that have ceased to pay dividends during the researched period will still be included in the study, because they might signal a decline in propensity to pay dividends, as observed by Denis and Osobov (2008).

The variables for quantitative study are internal factors:

• Dividend payout for the 10 year period under review.
• Earnings per share (EPS) growth rate for the period.
• Compound annual growth rate of sales per share over the period.
• Profitability of the company measured as Net Income divided by Total Assets
• Company size as measured by market capitalisation on the JSE.
• Total debt to equity ratio will be used to measure leverage and
• Liquidity will be measured by calculating the Current Ratio (CR)

Since the secondary objective of the study is to find out the external determinants of dividend payout decisions by finance executives, a survey method was employed to establish attitudes to those factors. The survey was conducted using self-administered questionnaire sent through to respondents via email. This survey method was used because self-administered questionnaires are time-saving and very cost effective (Salkind 2012).

A tried, tested, reputable and reliable survey instrument, originally developed by Baker for Harvard University was chosen for the current research. This questionnaire has been adapted by Firer et al (2008) to the South African context. The adapted version was then sent to one financial director as a pilot to test the validity of questions, objectivity and to assess possible response time for the questionnaire. This instrument was chosen as appropriate for this study because it has been proven effective by Baker et al (2009) in their survey of Indonesian executives, and by Firer et al (2008) when they studied dividend policy in South Africa.

In further tweaking the questionnaire, the variables that have both positive and negative impact on dividends were selected. In order to understand company issues, JSE and company annual reports were reviewed and used as a guide in the adaptation of the questionnaire. Questions were drafted close-ended, phrased as statements using 5 point Likert continuum scale to assess degree of dis/agreement. Responses were then graded out of 100% to compute mean scores from the respondents’ answers. E.g. a score 4/5 can be quantified to be 80% and vice versa. The scores in % form would then be tabulated alongside the random sampling numbers and various subgroup categories for data analysis. As the written authorisation for the use of Baker questionnaire is still not present, therefore  – to avoid plagiarism –  only the draft questionnaire with sample questions is attached to be reviewed by the lecturers.

Measures are in place to ensure that respondents are given sufficient time of 45 days to respond. Automated reminders; follow-up emails and phone calls will be made at 10 day interval to ensure that maximum response rate. This will help the researcher maintain an adequate sample size and the integrity and representativity of the results.

An introduction letter indicating support from the University of South Africa, objective of the research, confidentiality statement, and the reasons why the specific respondents were chosen to partake in the study will be sent via email together with the questionnaires to all respondents.

Data analysis

For this study the dividend payout is the dependent variable that will be examined in connection with the various internal and external factors, including company size, company growth, profitability and macroeconomics (tax, share price etc.), being an independent variable in statistical analysis. Prior to commencement of any data analysis, the collected data would be verified, sorted, properly tabulated and grouped according to various subgroups. Any missing or dubious data will be cleaned up.

The data will be analysed three-ways, using both descriptive and inferential statistical methods. Firstly, descriptive statistics will be used to describe company financial performance, and the means and standard deviations will be computed to measure central tendency and variability, respectively for all data sets collected (measuring dividends, growth, profitability, liquidity, leverage and size).

Secondly, to determine if there is a relationship between internal factors and dividend payout, Pearson product moment r correlation coefficient will be determined. Thirdly, t-tests and analysis of variance will be computed to establish the significance of any relationship that has been established.

Thus, the following null hypotheses will be tested:

1. There is no relationship between company size and dividend payout as measured by mean scores of market capitalisation and dividend per shares payout.
2. There is no relationship between company growth and dividend payout as measured by mean scores for sales growth, EPS growth and dividend per share payout.
3. There is no relationship between profitability and dividend payout as measured by mean scores for Net Income/Total Assets and dividend per share payout.
4. There is no relationship between liquidity and dividend payout as measured by mean scores for Current Ratio and dividend per share payout.
5. There is no relationship between leverage and dividend payout as measured by mean scores for total debt to equity ratio and dividend per share payout.

Lastly, data collected from the questionnaires will be analysed using Pearson r coefficient product, to ascertain that there is a correlation between the variables (e.g. tax rates, share price, etc.). If correlation exists, it will then be analysed further using one tail t-test to determine if it can be used for generalization purposes. Confidence or significance levels of 95% will be used for all statistical analyses conducted during this investigation. This method of data analysis has been used before by established researchers (Myers & Bacon 2004;  Baker & Powell 2012) in their various studies of dividend policy.

Limitations of the study

The authors anticipate challenges and limitations during data collection and fieldwork that might affect the outcome of the study. During the period under review (2002 to 2012), the mining industry experienced a boom, and global markets crashed in 2008. Thus, with industrial decline, encountering insufficient data for companies affected by delisting and merger activities is likely. Newly listed companies might also have limited or insufficient data. Furthermore, some companies might have delisted from JSE due to merger and acquisition activity. Due to mining being a long term capital intensive investment, some of the companies under review might still be in their development phase, and not yet paying dividends.

Another challenge could be the availability of finance executives, because of the congested itineraries that they usually have. This might result in late submission of questionnaires, rushed submissions or non-submission at all and might have an overall impact on the research with longer than anticipated research time. To eliminate this bias the respondents will be contact via email and telephone. The follow-up will be costly, and due to limited financial resources available for this study, will only be done in 10-day intervals..

Ethical considerations

As the researcher, I will take responsibility for this study and ensure that it is carried out in an ethical manner. Due consideration will be taken to ensure that the objectives of this study are explained in languages understood by all participants. The respondents will be requested to sign an informed consent form to confirm that accordingly. They will also be assured by the researcher that confidentiality will be maintained throughout this study.

Furthermore, to guarantee privacy, anonymity will be applied by the use of random numbers instead of company names. The participants will also have an option to opt out if at any stage if they feel uncomfortable to proceed further. Participants will be given an opportunity to comment on the provisional report prior to publication of the final report. Their comments will be taken into consideration and be incorporated into the final report. The results published will reflect the group analysis and not necessarily individual company performance.

Reference List

Allen, F. and Michaely, R. 2002. “Payout Policy”. North Holland Handbook of Economics.

Ardestani, H. S, Rasid, S. Z. A., Basiruddin, R. and Mehri, M. 2013. “Dividend payout policy, investment opportunity set and corporate financing in the industrial products sector of Malaysia”. Journal of Applied Finance and Banking, 3 (1): 123-136.

Baker, H. K. and Powell, G. E. 2012. “Dividend policy in Indonesia: survey evidence from executives”. Journal of Asia Business Studies, 6 (1):79-92.

Creswell, J. 2003. Research Design. 2nd Ed. London: SAGE Publications.

Denis, D.J. and Osobov, I. 2008. “Why do firms pay dividends? International evidence on the determinants of dividend policy.” Journal of Financial Economics, 89 (2):62-82.

Ferreira-Marques, C. 2013. “Glencore Xstrata promises dividends, buyback, says it will beat synergy target”. BusinessDay. 03 May 2014.

Firer, C., Gilbert, E. and Maythan, A. 2008. “Dividend Policy in South Africa”. Investment Analyst Journal, 68 (1): 1-20.

Fraser, C. 2012. “The ins and outs of dividend paying mining stocks”. Resource Investing News. 14 May 2012.

Gitman, L.J. 2009. Principles of Managerial Finance. 12th ed. London: Pearson Prentice Hall. Global Competitiveness Index Report 2013-2014 edition.

Gravetter, F. J and Forzano, L. B. 2009. Research Methods for the Behavioural Sciences. 3rd ed. Belmont: Wadsworth Cengage Learning.

https://www.boundless.com/finance/textbooks/boundless-finance-textbook/dividends-15/what-are-dividends-113/dividend-irrelevance-theory-477-8290.asp  Accessed on 19 January 2015.

http://www.investopedia.com/terms/d/dividend.asp Accessed on 20 August 2013.

Kawano, L. 2014. The Dividend Clientele Hypothesis: Evidence from the 2003 Tax Act. Office of Tax Analysis Working Paper 102. American Economic Journal: Economic Policy,  February 2014.

Kotze, T. 2007. Guidelines on writing a first quantitative academic article. 2nd Ed. Pretoria: University of Pretoria

Malik, F., Gul, S., Khan, M. T., Rehman, S. U. and Khan, M. 2013. “Factors influencing corporate dividend payout decisions of financial and non-financial firms”. Research Journal of Finance and Accounting, 4 (1): 35-46.

Marks J. 2001. “Empirical findings on Director’s views on dividend policy” Management Dynamics, 10(2): 50-66

Myers, M. and Bacon, F. 2004. “The determinants of corporate dividend policy”. Academy of Accounting and Financial Studies Journal, 8 (3):17-28.

Nell, A.E., Hamman, W. D. and Smit, E. 2001. “Dividendveranderings in Suid Africa-Tekenend van die verlede of toekoms?” Studies in Economics and Econometrics, 2 (2):50-66

Porta, R., Lopez-de-Silanes, F., Shleifer, A. and Vishny, R. W. 2000. “Agency problems and dividend policies around the world”. The Journal of Finance, 15(1):1-33

Salkind, N. J. 2012. Exploring Research, 8^th ed. New Jersey: Pearson Education.

Venter, P. 2013. Tutorial letter 102: Study guide for HPBUS81. Pretoria: UNISA.

Verhoef. M. J. and Hilsden, R. J. 2001. Writing an effective research proposal. Calgary: University of Calgary.