Print Reference This Reddit This

Dividend Payout And The Effects Of Firm Characteristics Finance Essay

One of the most fascinating topics in Finance literature both in theoretical and empirical dimensions is the dividend policy of the firm. Extensive research has been done in many facets of this topic. First the dividend irrelevance proponents challenged the norm that dividend payouts impact the value of the firm, contending that there is no relationship between the dividend policy of the firm and the value of its shares. Miller and Modigliani (1961) were first to propound this theory under the assumptions of perfect capital markets. Thereafter most literature on dividend payouts have taken root from the standpoint of imperfect capital markets .The generational “bird in hand view” (Gordon (1962),Lintner (1962) came as a counterview to the Modigliani and Miller (1961) theory that investors are indifferent as to whether their returns arise from dividends or capital gains. They contend that investors care more about dividend paying stocks than non-dividend paying stocks and as a result of this underlying interest the market prices of such stocks will be higher, thus enhancing the value of the firm. Several schools of thought appropriately tagged dividend theories have emerged subsequently with respect to whether dividend policy of the firm is relevant or not in determining its value. Notable among the theories are the dividend signalling, agency cost, residual, tax clientele, free cash flow and pecking order theories. Therefore against the background that dividend is irrelevant, one wonders why dividend is paid, that is, why do they care about their dividend policy (the dividend puzzle: Black, 1976)? At the other extreme the supporters of dividend relevance believe that the answers to the puzzle can be explained by the different theories mentioned above. The general norm is that dividend is invariably paid and if so what determines it may vary among firms or countries or indices or over periods.

The discourse on the dividend in the decade has been dominated by the extent to which the changing characteristics of the firm influence the likelihood of dividend payment. Fama and French (2001) report the consequence of changing characteristics of the firm and the lower propensity to pay. Baker and Wurgler (2004) also find a decline in number of dividend paying firms. Allen and Michaely (2003) find a decrease in the number of dividend increases. In their own investigation, Gruillon and Michaely (2002) document a gradual substitution of share repurchases for dividends. On the other hand DeAngelo et al (2004) through a sector-specific study of industrial firms find the incidence of dividend increase. On the emerging market scene, Reddy and Rath (2005) find the effects of changing characteristics of the firm in India. In the UK, Trojanowski and Renneboog (2005) find that dividends still constitute a major proportion of the distribution among firms in the U.K. market that shares the same attributes of active corporate control, liquidity, diffuse ownership, common law system and strong share holder protection.(La Porta et al (2000)as cited by Trojanowski and Renneboog (2005))

The above scenarios underpin the importance of this study. Motivated by these, this study notes that the content of literature pertaining to the changing characteristics of the firm and their impact on dividend payout in the decade has been mostly in the US market. Though studies exist on the UK market which shares similar market attributes, it is not as wholesome as the US market and the degree of unison has not been converging enough. Are there differences on the influence of the firm characteristics in different markets over different time periods? This study will add to the literature on the UK market.

Most importantly also, the study recognises the importance of the intervening period between 2005 to 2010 during which the financial crisis occurred (2007/08) and attempts to determine the magnitude of influence the characteristics of the firm had on their likelihood of dividend payment over the pre-crisis period, post crisis period and the entire study period. The relationships between the firm characteristics and the likelihood of dividend payments among UK firms will be hypothesized over the three periods and inferences will be made through empirical analysis. Fama and French (2001) adopted the firm characteristics of size, profitability and investment opportunities as the explanatory variables with indicator ratios and this would also be used in the is research on the UK market. Through this study the characteristics of the firms and the patterns of dividend payments over the periods will also be assessed and this will be useful for further comparisons.


Firms distribute value to shareholders through five major ways: regular cash dividend, open market repurchases intra-firm tender offers, targeted repurchases and special dividends (Barklay and Smith 1988). Dividend is therefore one of them and the most puzzling (Black 1976). Fama and French (2001) note that dividend has long been an enigma and the fact that it is being paid remains difficult to explain.


Modigliani and Miller (1961) were the “authors” of the groundswell proposition on dividend irrelevance. They assume perfect capital market devoid of taxes, transaction costs and any form of information asymmetry. However it does appear that in real world those restrictive assumptions that underlie this theory cannot hold, hence the market Imperfections and rightists school of thought assert that since imperfect capital market conditions prevail, what matters more is not whether dividend is relevant but what are the factors that determine it. However on the face of it, there is no claim on the part of Modigliani and Miller that shareholders do not desire dividends; the assertion is that where market is perfect, the dividend policy does not create value above what the investment policy can generate (DeAngelo et al 2009).


Most literature that built up after the irrelevance proposition are focused on different theories that include dividend signalling, agency cost, tax clientele, transaction cost, residual and pecking order theory. They were all based on imperfect capital market conditions


Finance literature is abound with studies on the various factors that are specific to the firm which determine their dividend payout .They include but not limited to business risk, agency cost, profitability, size, institutional holding, insider ownership, tax clientele, leverage, free cash flow, investment growth opportunities, leverage etc. The various factors are easily subsumed within the theories.

The influence of agency cost is also often discussed and the general consensus is that dividend mitigates agency cost since managers will not have excess cash to “throw about” or engage in “empire building” ( Jensen et al(1992), Rozef(1982). In the same vein , Ang et al (2000), through their organisational study draw empirical evidence that higher managerial ownership reduces agency cost. Rozef (1982) also buttresses the inverse relationship that would ordinarily exist between low managerial ownership and dividend payout as well as the trade –off between transaction and agency costs. If firms’ dividends are used to avoid agency cost, investors must be mindful of the costs of raising funds in case an investment opportunity presents itself. Coerced regular distributions means that external financing must be resorted to for profitable investments; this comes with a cost ( Eastbrook 1984) . Similarly Brockman(2009) comments that agency cost of debt appears more important in the determination of dividend policy. Allen and Michahely (2003) however differ with models that assume that managers can only be influenced to pay dividend but cannot be disciplined to avoid negative net present value investments. Under this ambit therefore insider ownership is recognised as a factor that can affect dividend policy. Suffice it to add that the extent of dispersal of ownership and indeed ownership structure become determining factors.

Another factor that takes a central theme is business risk. Researchers have defined it as the element of uncertainty that the future profitability of the firm might deviate from the current state. The greater the uncertainty, the greater the implied business risk and therefore the lower the dividend payout ( Jensen et al 1992); (Al-Kawuri 1993). Expectedly it is argued that firms designated as high risk command high volatility of stock prices than low-risk counterparts. The debate on the business risk of the firm cannot be complete without mention of the dividend signalling action of the firm with which firms carry information about the future of the firm.

While it is easy to conclude that dividend signalling and its direction can predict the future, Miller and Rock(1985), Lintner ( 1956) adduce that dividend increase are mostly accompanied with future certainty of continuity. It therefore has informational content. The dividend announcement impact study of Alharoy and Swary (1980) also lays credence to the signalling power. It is however noteworthy that this predictive power of signalling is not without opposition in finance literature. Amihud and Kefei(2006) suggest (in their investigative work on the declining information content of dividends) that if dividend increase announcements are becoming less informative , such dividend-related costs should be avoided. As a corollary, Stein (1979) argues on the costliness of signalling by firms while DeAngelo et al (1996) analyse large dividend paying firms and find no special effect of signalling. In a nutshell signalling has a role to play in determining the business risk of a firm even as the issue of the extent of its predictive power remains matter for debate.

The tax clientele of the firm has also occupied a notable place in literature on the determinants of dividend policy. It is believed that since various classes of investors are taxed differently, those in low tax bracket would opt for high dividend paying stocks and vice versa for high tax payers. Litzenbeger and Ramaswamy (1980) and Brennan(1970) document that high tax investors prefer capital gains because of the the tax disadvantage. Thus the tax clientele argument introduces the factor of institutional holding effect depending on the side of the bracket they fall in, because of the shareholding influence – Amihud (2006),Short et al (2002) and Khan(2006) provide proofs . It is also assumed that dividend payments will be less preferable to firms since its payment places them at a competitive disadvantage in terms cost of raising funds through external financing. Fama and Fench (2001) recognise that such firms will have higher cost of equity than those that do not pay dividends. Therefore a good question is why the firms pay dividends even when they have a high concentration of high tax bracket investors. For instance Allen and Michaely( 2003) report that the holding of dividend paying stocks in US is high among investors in the high tax bracket. In summary what makes tax clientele an interesting factor in payout consideration is that even though capital gains appears a better choice ( as a tax avoidance mechanism), dividend is seen as possessing benefits that are capable of offsetting its tax disadvantages.(Mollar 2001)

The impact of transaction cost on dividend policy of the firm cannot be overemphasised. It is mostly viewed as the cost of acquiring external fund. When this is high, the trade off will be on lower dividend payout. On the contrary higher dividends will mean that transaction costs will invariably be higher since funding cannot be attained through retained earnings (Fama 1974). Furthermore dividend policies that address the issue of external funding also connect with the residual and pecking Order theories. From the residuals point of view the ability of the firm to distribute dividends depends on the extent of its need for profitable investments (Saxena 1999).This investment need that will potentially trigger the need for external financing is also where the pecking order comes in since firms are expected to behave in the order of exhausting internal financing, utilising debt before resorting to equity issue (Donaldson (1961).

A usual extension of the discourse on transaction costs brings up the effect of leverage. This is recognised as carrying high costs which limits dividend that a firm can pay. Expectedly, it is viewed as a strong determinant since the firms are not in a position to pay dividends without even acquiring more financing costs Jensen et al (1992), Mollar (2001).

The size of the firm is also as important and most experts believe that it offers benefits of reduced cost of financing and operations (because of economies of scale). The larger the size therefore, the better the chances of paying dividend. Redding (1997) finds that larger firms are in a better position to distribute dividends than the smaller ones. On the other hand, Holder et al (1998) discover that this latitude enjoyed by larger firms comes from easier and cheaper access to funds.

Perhaps some of the most often discussed determinants of dividend policy are profitability and investment opportunities. Fama and French (2001) assessed the effects of these characteristics in addition to size on the probability of dividend payments. Profitability is affected by transaction cost and so even as traditional literature recognises a positive relationship between profitability and dividend, it also admits an inverse relationship between profitability and costs (Han et al 1999). In relation to investment opportunities, most firms that have high growth opportunities prefer using internal funding which impairs their dividend payment propensity (La porta et al 2000).The utilisation of internal funds for growth opportunities is often seen as a control for the free cash flow of a firm such that managers are discouraged from frivolous activities that may be induced by excess cash. The usual belief is that there is a positive relationship between free cash flow as a factor in dividend payouts and this relationship is used to control the agency cost of free cash flow.

The subject of dividend determinant has also been discussed in the context of its relationship with share repurchase as an alternative payout channel in terms of possible substitution. Fama and French (2001) explain the declining phenomenon as a function of the changing characteristics of the firm. Some studies on the UK market however document that dividend still accounts for large portion of payout even with increasing role of share repurchases (Trojanowsli and Renneboog 2005). On the other hand Grullon and Michaely (2002) find that the increasing role of share repurchases contribute to the diminishing dividend. Depending on the base of study the outcome can only be empirically justified.

In conclusion, the above represents the different views on the relevance or otherwise of dividends and points out the determinants of dividend payments and the underlying theories which further support the influence of the factors on dividend payouts.


The motive of the study is to find out the effects of firm characteristics on their likelihood of dividend payments, with UK market as the evidence.

The hypotheses will be tested to find to find whether any significant relationship exists between the probability of dividend payment and the firm characteristics.

These will be tested for the three periods –pre-crisis, post-crisis and for the entire span of study and the following questions will be answered.

Is there any relationship between the variables and the probability of paying dividend over time? Specifically for the whole period of study and the periods pre and post 2007/08 financial crisis.

What are the possible differences in the relationships that are distinct to the respective periods of the study

How do the variables alter over time under different scenarios

What is the pattern in dividend payments as a whole and among the different dividend class .The dividend classes will be designated as Payers, Non Payers, Never Paid, and New List( Fama and French 2001).

How do the respective dividend policies that dwell on the various firm characteristics find support from this study?


Drawing from the background of the research and the questions raised, the following are the objectives of the study:

To establish any relationships between the firm characteristics and the probability of dividend payments over the phased periods of the study

To identify possible differences in the relationships exhibited over the different periods.

To segregate the firms along classes with a view to carrying out a more in-depth analysis of their characteristics and patterns

To ascertain whether there is any period –specific influence of the variables on dividend payout among the firms. Specifically what has happened over the pre-crisis, post-crisis and the entire sample period.

To answer the question as to how the notable dividend theories that relate to the identified proxies hold in the face of empirical evidence and in the market of study



This study investigates the relationship between the firm characteristics and the likelihood of paying dividends. Different studies have been carried out in literature using different characteristics that cut across profitability, size, ownership structure, investment opportunities, institutional holdings etc. Fama and French (2001) identified and used size, profitability and investment opportunities in their groundswell finding. Others have also investigated this causal effect in terms of proofs to the dividend theories (Alli et al 1993), Mollar 2001)

The sample for the study is the list of firms in the FTSE 100 INDEX. The index is chosen due to its relative importance in the ranking of indices around the world and also representing the top 100 most capitalised firms in the UK.

The sample period will cover 2005 through 2010. In order to estimate the relationship between the probability of dividend payment and the variables, the study and test of the hypotheses will be done in three phases—Pre-crisis period, post-crisis period and the entire period. This differentiation will also help to ascertain the possibility of period effects.

Furthermore to assess the effects of these identified characteristics on the dividend action of the firms , they are divided into different subgroup thus: All firms(including payers and non payers), Dividend payers(paid dividend at end of financial year t ), Non Dividend Payers( did not pay dividend at end of financial year t), Former payers (paid in year t-1), Never paid( did not pay dividend between time t-1 and t), New List ( listed on the exchange during time, t) Fama and French (2001).

5 2 DATA

The secondary data for the research will be obtained mostly from the Thomson One Banker database which contains most of the accounting and market information of the companies listed on major stock indices. The data obtained will also be used to extract the relevant ratios accordingly. The database and the annual report of the companies are expected to satisfy the secondary data requirement of the research


The study identifies three firm specific variables of profitability, size and investment opportunities as the explanatory variables. These will be used to establish the relationship (if any) between them and the probability of dividend payment. Fama and French (2001)) adopted this approach. The hypotheses will thus be tested for the effect of each characteristic on the probability of dividend payments and over the phases of the study.

The different variables will be defined as follows with their indicator ratios





1. /

2. /

Ratio of aggregate earnings before interest for yr t to aggregate assets at end of year t

Ratio of aggregate after tax earnings to Book value of equity at end of year t

Investment Opportunities



Ratio of market value to book value of assets at end of year t

Ratio of change in assets between year t-1 and year t



2. /

Aggregate assets

Ratio of aggregate liability to aggregate assets

Source: Fama and French (2001)


The study will first use descriptive statistics to present and interpret the patterns and characteristics among the dividend classes (payers, Non Payers, Never Paid, New List) and across the years of the study (Fama and French 2001)

The Logit regression model will be used to determine the relationship between the firm factors and the probability of dividend payments.

The T-test will be used to test the hypotheses and ascertain the strength of the relationship and likelihood of paying dividend.

The Logit model will be:


Represents the probability of dividend payment at time t. It equals 1 if dividend is paid and 0 if not paid

Exp represents the exponential function

, …. represent the independent variables

, …….. Represent the regression coefficients




Research proposal

29th April 2011


1st week in June

Literature Review

8th July, 2011

Data Collection

22nd July, 2011

Data Analysis, Interpretation and Findings

10th August, 2011

Final Draft

20th August,2011

Proof Reading, Corrections and Improvements

25TH August, 2011

Formatting, Printing and Binding

28th August, 2011


31th August, 2011


The resources for the research in terms of data will be mostly sourced from the University of Leeds web platform to collect the relevant data for analysis and interpretation. It has provides access to renowned data bases that will aid the research. Other research needs will be accessed from the websites of the companies if required.

The project will also require some financial commitment in the process of its execution and a sum of £250.00 has been budgeted for the process

Need help with your literature review?

Our qualified researchers are here to help. Click on the button below to find out more:

Literature Review Service

Related Content

In addition to the example literature review above we also have a range of free study materials to help you with your own dissertation: