The dividend policy of firms has long been recognised as an enigma for academicians since the ideal perfect capital market theory suggests that there should not be any value-adding role of dividends in relation to the market value of firms. Again, however, there is much evidence to suggest that capital markets are not efficient, and that dividend payouts do matter as far as the value of a firm is concerned.
This study will focus on the various studies carried out in the arena of dividend payout theories, such as Irrelevance Theory, Lifecycle Theory and Signalling Theory, as well as the empirical findings relating to these. More specifically, the study focuses on the bank dividend payout policies of Saudi Arabian Islamic banks. The reason for choosing the banking sector is that, although dividend policies have been studied in-depth for non-banking firms, thus far, an inadequate number of studies have been centred on banking firms. In addition to this, there is also an interesting opportunity to study Islamic banking practices, which are significantly different from commercial banks. It is expected that the study will shed light on a so far unexplored area.
2.2 Theoretical Background
2.2.1 Dividend Irrelevance Hypothesis
Dividend irrelevance hypothesis was first introduced by Miller and Modigliani (1958, 1961), the main argument of which is based on efficiency market hypothesis, which holds that, given no information asymmetry between outside shareholders and the insider managers of firms, in addition to no transaction costs, the value of the firm is independent of the financial structure of the firm; inother words, whether the firm is equity-financed or debt-financed, or whether or not the firm is announcing dividends, is irrelevant; the only matter under consideration is whether or not the firm is investing in positive NPV projects. Hence, according to this hypothesis, dividend pay is irrelevant to the valuation of the firm in the market.
Again according to the above theory, if a firm does not pay dividend then the shareholder will receive an amount equivalent to gain in terms of price appreciation, and by the same logic, if the firm pays dividend, there will then be an equivalent amount of price drop (referred to as EX dividend cut); hence, shareholders' wealth remains constant. Along with this, shareholders can also manufacture so-called home-made dividends: for example, if some shareholders want dividends and others do not, and then they simply trade and satisfy their needs.
In reality, however, dividend pay is viewed as being relevant since there are imperfections in capital markets, such as information-asymmetry problems; thus, many theoretical models have been presented as suggesting that managers make dividend policies in order to maintain the market value of the firms (Black, 1976); Allen and Michaely, 2002). Dividend signalling theories have also proposed that firms pay dividends in order to signal the future financial health of them; again, this might help investors to distinguish between good and bad firms, thus helping to solve the problem of adverse selection.
2.2.2 Information Signalling Theory
In reality, capital markets are not perfect, with the most common imperfection recognised as the information asymmetry problem, i.e. the difference in information between the insider managers of a firm and outsider shareholders. Due to this difference in information, there can be two main types of problem: adverse selection and moral hazard. Adverse selection is an ex ante problem, which occurswhen the firm decides to raise capital in the share market. Importantly, if the investors do not have enough information to distinguish between bad and good firms, an average pricing is placed on both, which might deter the good firms to remain in the market. Moral hazard is an ex post problem, and is related to investment decisions that are made by managers but which are not known to outsiders; hence, there is always a chance that shareholders' wealth will be mismanaged (Hail et al. 2012).
In both of these cases dividends can be used to provide a solution, such asin the case of the Dividend Signalling Theory, which suggests that dividends be used as signalling by good firms since it cannot be copied easily by bad firms since they might find it difficult to continue paying dividends in future; hence, this might help investors to distinguish between good and bad, and thus solve the adverse selection problem (Tsuji 2012).
Information signalling theory suggests that managers do pay out dividends as information signals. Specifically, good firms can use dividends as signalling future profitability. Importantly, it is difficult for bad firms to emulate good firms' signalling since they will not be able to remain committed to paying dividends in the future; hence, dividend payout can help to solve theadverse selection problem(Tsuji 2012).
Dividends can also be used to solve the moral hazard problem since the Dividend Lifecycle Theory argues that it is always better for matured firms with shrinking investment opportunities to pay dividends so that shareholders have confidence that their wealth is not destroyed; in other words, this helps to mitigate agency conflicts (Tsuji 2012).
2.2.3 Free Cash Flow Hypothesis
Jensen (1986), on the other hand, proposes a theory that is widely known as the Free Cash Flow Hypothesis.This study was the first formal investigation of the so-called agency conflict between outside owner shareholders and firm insiders who have a better information advantage and controlling power (in this context, the term 'control' refers to the control over free cash flow generated by the business), with the argument holding that the firm has to make the decision of whether or not to plough back the free cash flow or to otherwise distribute it as dividend gain to shareholders(Hyderabad 2013).
Accordingly, if firm managers invest in positive NPV projects, shareholders' wealth is created; otherwise, this might be destroyed by managers if they invest in riskier projects or, in some other way, gamble with shareholders' wealth. Hence, there is importance associated with dividend payout in this regard, i.e. dividend payouts might put constrains on managers, stopping them from investing in bad projects(Karpavicius and Yu 2012). This suggestion is further supported by the empirical evidence on the Dividend Lifecycle Theory (De Angelo et al. (2005 please check).
2.2.4 Tax Effect Hypothesis
The MM theory, as highlighted previously, was based on the assumptions of no transaction costs and no taxation. Realistically, however, taxes are charged on both capital gains and dividends; however, in many countries, dividends are charged at a higher rate than capital gains, hence giving the incentive to rational investors to demand capital gain more so than dividends (Carroll and Prante 2012). [1]
Moreover, the capital gain tax is levied only when shares are sold, whereas dividend taxes, on the other hand, are mandatory. Accordingly, from this point of view, rational shareholders are better off adopting the first scenario (Kenchington 2013).
2.2.5 Agency Cost Theory
The modern corporate finance is centred on the principal agent conflict, which means that the principal owners of the firms are not the insiders. In this case, there is always a possibility that the insiders do not act in order to maximise shareholders' wealth. Again, insider managers might use their superior information and control to mismanage shareholders' wealth and increase their own personal wealth, such as through the application of bonus schemes and employee stock options (Ganguli and Chaturvedi 2012).
Hence, the dividends payouts gain importance because regular dividend payouts put a constraint on managers to mismanage shareholders' wealth, which is more relevant for those firmsthat are large and matured since they already have fewer investment opportunities. With this noted, it can be stated that such firms should pay dividends inorder to preserve shareholders' wealth.
2.2.6 Clientele and Substitution Effect
Modigliani and Miller (1963) claim that there can be clientele of investors related to dividends: for example, there may be investors whose immediate need is a steady dividend, whereas there may be others who need capital appreciation more so than dividends. However, the same authors also suggest that investors can generate home-made dividends and trade with each other to satisfy their specific needs, thus meaning the impact of dividend payment overall is negligible.
Brav et al. (2005)state that retail investors prefer cash dividends more so than when there are no tax benefits. The same author interviewed the managers and found that most of them suggested that retail investors consider dividends to be really important, and that preferences for dividends grow with age.
2.2.7 The Bird in the Hand Theory
Alzomaiaand Al-Khadhiri (2013)
This theory criticises the MM argument, and further holds that investors prefer dividends since, with the increase in dividend payouts, the risk for holding stocks decreases. In this regard, the argument is contrasting;since capital gain is riskier compared dividends, it is good for firms to pay dividends and preserve shareholders' value (Gordon, 1963;Lintner,1962).
2.2.8 Financial Lifecycle Model
It is important to review latter developments, such asthe Lifecycle Model, according to this theory owing to the fact that dividend payout patterns change over the lifecycle of firms. For instance, firms that are in the growth stage would seek to retain free cash flow and invest, whereas matured firms may have fewer investment opportunities.
De Angelo et al. (2005)observe that, even on a cross-country basis, dividend-paying firms are mature and large (as indicated previously by the Agency Cost Model/ Free Cash Flow Problem Model above), which is related directly to the fewer investment opportunity available. More in-depth, the model is based on the relative cost benefit, like the cost is the agency cost involved (compatible with the Agency Cost Model described above), whereas the benefit is related to floatation costs/costs of external debt (since retained earnings are available for reinvestments), the cost-benefit trade-off changes over the lifecycle of firms, as found byFamaand French(2001), Grullon(2002), De-Angelo and De Angelo (2006), amongst others.
It seems that dividend-signalling theories are not capable of explaining why matured firms pay more dividends since they already have less investment opportunities; hence, dividends cannot be used as the signal of future profitability, and as such, the agency conflict perspective seems more relevant.
The very study uses RE(retained earnings)/TA(book value of assets) and RE/TE(book value of equity)as the proxy variables for the lifecycle stage, and argues that this so-called ' retained earnings to contributed capital' mix signals firm maturity owing to the fact that, if the ratio is comparatively low, it might signal that the firm is still in the 'capital infusion' stage, and may therefore require more external capital to finance projects, whereas a comparatively high value means that there is enough cumulative profit that the firm can use to finance projects,i.e. self-finance; this is characteristic of matured firms. This paper convincingly drives home the point that, when controlled for factors such as firm size, current and recent profitability, growth andleverage, there is a significant positive association between dividend payout decision and the ratios.
2.3 Empirical Evidence
2.3.1 Dividend Irrelevance Hypothesis Empirical Evidence
Early empirical studies focus on the dividend irrelevance theorem. One of the pioneers in the field is the study by Black and Scholes (1974), during which theauthors studied 25 portfolios combining the high-dividend yield and low-dividend yield stocks, and analysed the correlation between stock returns and dividend yield. The results clearly showed no significant difference in stock returns between the two classes; in other words, dividend yields are not significantly correlated to stock returns. This result was the same both before and after taxation. Hence, this was an empirical support to the dividend irrelevance theorem; in other words, the market value of a firm is independent of dividend yield.
Recent evidence from the UK goes against the Irrelevance Theory. Dhanani (2005) deployed a survey method to explore the dividend policies of UK firms across sectors. This was a more exploratory study based on detailed questionnaire sent to managers across industries. The results show support for the dividend relevance hypothesis.
During more recent times, there have also been studies carried out in Asian economies. For instance, the work of Conroy et al. (2000) examined the impacts of expected dividends and earnings of the following year on the dividend announcements of the current period. However, this study-similar to many other event studies-showed that earnings announcements are more important in terms of explaining stock price behaviour rather than dividends announcements. Dividend announcements by firm managers have an insignificant impact on stock price movements.
The general consensus is that, under imperfect market conditions, the dividend Irrelevance Theory does not hold. Overall, there is always mixed support for the Irrelevance Theory, with earlier empirical studies showing some support; however, more recent studies find little support for the same.
(Uddinand Chowdhury(2005)studied 137 companies enlisted on the Dhaka Stock exchange with the aim of analysing any relationship between dividend payout and stock price movements. The study showed the opposite impact on stock returns: for example, ranging from 30 days before to 30 days after the dividend announcements, falls in stock prices were witnessed, which gave rise to 20% loss in shareholder value. Thus, the scholars suggested that current dividend yield can reimburse the diminished value to some extent; hence, generally, this study supported the Irrelevance Theory.
2.3.2 Free Cash Flow Empirical Evidence
In addition, Brav et al. (2005) claims that the corporate managers consider the dividends as of same importance as the investment decisions.This study lends support to the dividend relevance theory. Lastly, Foley (2007) documents that large US firms prefer to hold large cash on their balance sheet to avoid taxation; hence, the FCF theory does not hold good for large MNCs.
2.3.3Tax Empirical evidence
In the recent past, scholars Hassettand Auerbach (2006) analysed the impact of the income tax relief programme on dividend payouts. The research stated that a significant impact was witnessed across the equity market as a result of tax cut, where dividend-paying organisations witnessed increments in stock prices in addition to dividend payment, as opposed to a decline in capital costs. In contrast, however, non-dividend-paying organisations witnessed declines in their finance costs, as well as an investment stimulus, which is regarded as being in-line with the dividend taxation hypothesis. Accordingly, it remains unclear whether cost of capital reduction is significantly correlated with dividend payouts.
2.3.4 Other Empirical analysis
Rahman, Amin and Siddikee (2012)studied the impact of various dividend announcements on stock price movements. The study utilises a number of different forms of event study method, such as CAAR/MAAR, and accordingly analyses the impact of announcements on stock price movements and finds insignificant results. Significant negative stock returns are found prior to the announcement of stock dividends; this might suggest the speculative nature of investors. Again during the post announcement period, positive gains are seen since rational investors realise the possibilities of abnormal gains; however, as far as the cash dividend impact is concerned, there is no significant result for companies listed under Dhaka Stock Exchange for 60 days following the announcement. AAR shows no significant abnormal returns for cash dividends; however, for the stock dividend announcement, there are significant returns in the range of 6%.
Zakaria, Muhammad and Zulkifli (2012)
Suleman et al. (2011) analysed the correlation between price volatility and dividend announcement in Pakistan, with the method adopted similar to studies mentionedearlier. The data was extracted from the Karachi Stock Exchange for the period 2005-2009. The technique utilised was multiple regression method, which showed a clear and positive significant relationship between dividend announcements and stock price volatility; however, a negative significant correlation between growth and dividend announcements was also evident.
Hussainey et al. (2011) performed similar regression analysis for studying the correlation betweenstock price volatility and dividends for 123 UK firms over a large period of time. Their work was based on that of Baskin (1989). Similar to Baskin (1989), the model was modified by adding firm-level and country-level control variables.
Consistent with Allen and Rachim(1996), Australia results (Hussainey et al., 2011) showed a significant negative correlation between stock price volatility and payout ratio. In addition, a significant negative correlation was also identified between price volatility and dividend yield. Overall, the payout ratio turned out to be the most important driver of stock price volatility, with the debt levels and firm size recognised as the most important of variables amongst control variables.
2.4 Brief History of Islamic Banking
This section will look into the history of Islamic banking, and will highlight the differences between commercial banking and Islamic banking. One of the striking differences is that Islamic banking is an interest rate-free banking system. The suggestion was made to devise and implement an interest-free banking framework centred on the key Islamic principles of profit-sharing, referred to as Musharakah, and cost-plus (Murabaha) (Gafoor, 1995).
The rise of Islamic banking began during the 1970s when the respective economies became more mature. As a result, several research studies were published, including those by Al-Araby (1967), Siddiqi (1961, 1969), al-Najjar (1972) and Al-Sadr (1974).In addition, a significant degree of attention was directed towards the field of Islamic Finance in media, with a number of important academic conferences held in London. As such, interest grew-not only in Islamic countries but also in developed nations (Gafoor, 1995).
As a consequence of the above activities, a number of large Islamic banks were established during the 1970s, with notable growthat a rate nearly of 10% perineum recognised by such firms, with their operations quickly expanded across borders(Hassoune, 2004). As a direct result of such wide-reaching developing, the expansion of the industry spanned the world, with the entire banking systems of a number of different countries, such as those of Iran and Sudan, changed completely in order to ensure adherence to the standards of the Shariah.The growth of Islamic finance has been driven by increasing demand by Muslim populations for Islamic financial products to be made available across the globe.
It would be an interesting study to analyse the dividend payout policies of Islamic banks, as well as how such policies impact shareholders' wealth.
2.5 Conclusion
Throughout this report, the key empirical researches and pivotal theories centred on dividend policy have been introduced and explained. Notably, dividend policy-related theories are described below:
Classical Dividend Irrelevance Hypothesis argues that, in the presence of ideal capital market conditions, a firm's dividend policies do not impact the market value of the same. Again, stockholders can create so-called homemade dividends and trade with each other.
The argument is based on the Agency Conflict Theory, which suggests that, although insider managers should act in order to maximise shareholders' wealth, since insider managers have more information advantage, they are in a position to gamble with shareholders'money; thus, paying out dividends might be one way of restricting the bad behaviours of managers.
Tax Effect Hypothesis argues that, since dividend gains are taxed at a higher rate than capital appreciation, it is much more rational for shareholders to prefer the second option. However, as has been established, this simple hypothesis is not very well supported by empirical evidence.
.
When reviewing the empirical and theoretical contributions in the literature, a number of gaps were seen, as discussed below:
Since most studies are focused on developed markets, emerging economies are not explored; hence, there is huge potential to survey the dividend policy impacts on shareholders' wealth in these areas.
Most researches are focused on non-financial and non-regulated firms (Partington, 1985).There are very few studies centred on regulated industry/banks; these studies show that dividend payouts by banks are different (Dickens et al., 2002).Again, more recent studies (Onali, 2009) suggest that dividend payouts by banks might also impact the risk of individual banks; this seems to be an entirely new area of research.
Another motivation for this work is that there is no important study carried out on the dividend policy of Islamic banks. Hence, this study might shed some light in this area and accordingly inspire further studies. Although Islamic banks are also regulated, regulatory rules are significantly different from commercial banks, meaning this opens up another opportunity to study the impact of dividends under a different regulatory environment altogether.
Hence, overall, it is expected that this study will help focus to be placed on a thus far unexplored area, although this would be challenging since there is minimal literature and studies available on the dividend policies of Islamic banks. The economy chosen is Saudi Arabia since it is recognised as being a fast-growing economy, whichhas also accepted financial liberalisation; hence, there is much competition in the banking sector,
