Interest rates differential and substantial transaction costs also interfere with the dividend irrelevancy argument based on perfect capital market, means that an individual is unable to 'costlessly' adjust his/her dividend pattern to fit his/her preferred consumption pattern. On the other hand, shareholders might give preference to the companies to provide them a set of desired dividend patterns, thereby creating a certain demand for specific patterns. The different dividend policies of the companies also attract investors to invest in, and when the policy changes, investors also adjust their stock holdings accordingly. This adjustment causes the movement of stock price. Unfortunately, it means that shareholders may incur costs of adjustment, but such costs may be avoided to the shareholders by an easy identifiable dividend pattern. In the same way, consequential costs may incur in the form of missed investment costs or opportunities of raising finance due to free cash flow shortage.
By keeping in view the entire described scenario, it is best in the company interest to follow consistent dividend policy that could minimise both consequential cost as well as adjustment costs for the investors. Investors are always risk averse and attracted towards companies that have high payouts and prefer regular cash income with relatively low tax bracket. Growth companies attract investors with no pressing needs for cash in high tax brackets and they normally pay lower dividends and reinvest more cash flows in expansion and new projects.
The need of information is another imperfection of capital market which is neither universally available nor costless. Therefore, the declaration of the dividend which is both universally available and free is thought to signal information to the market. Due to the reduction in the dividend is often read by the clientele as unfavourable information; managers are always reluctant to cut the dividends and the only increase the dividend if they are sure and that future free cash flows will enable them to retain the establishment pattern.
In 1994, an American pioneer company in strategic move, Florida Power & Light (FPL) cut its dividend by 33% and as predicted stock price immediately fell by 15%. The utility index of FPL's stock in two weeks time is show below.
[Figure 3-1]
After the announcement of dividend policy, FPL was under pressure that they would not be able to maintain the payout average of 90% for the best interest of shareholders. To avoid current pressure management announced that the policy will be reviewed in the beginning of 1995 instead of middle of the year. At the same time, United States changed the tax code to made capital gains more attractive than dividends. After that analysts came to the conclusion that FPL's action was a strategic decision to strengthen company's prospects for growth instead of signal of financial distress. This tactic gained new ground in limiting clientele effect. (Chew, 1993)
3.4.5 Life Cycle Theory
DeAngelo, DeAngelo, (2006) noticed that "Dividends tend to be made by mature established firms, plausibly reflecting a financial life cycle in which young firms face relatively abundant investment opportunities with the limited resources, so that retention dominates distribution, whereas mature firms are better candidates to pay dividends because they have higher profitability and fewer attractive investment opportunities."
quoted by Maria R. Borges in "Is the dividend puzzle solved?" 2008.
She also mentioned that dividend lifecycle expectations rely on the trade-off between the advantages and disadvantages of dividend payouts and it grows with the maturity and profitability of the firm. New firms need funds and resources in the beginning and because of this reason they do not pay dividends and also preserve internal resources to avail opportunities. On the other hand, big firms pay a small percentage of their profits as dividend and hold rest of the earnings to finance continuing growth and investment opportunities. Fama and French (2001) agreed with DeAngelo and DeAngelo's theory of lifecycle and found that high profit/low growth firms likes to pay dividends but low profits/high growth firms likes to retained their profits rather then paying dividends.
Borges (2008) mentioned that the lifecycle theory of dividend is comparable with outcome model of La Porta et al. (2000) which states that firms with higher investment opportunities have lower payouts because investors are aware from current situation of the firm and they do not exert pressure. La Posta et al. and some other practitioners confirmed this prediction through empirical analysis.
Lifecycle and agency theories are considered to be best positive theories of dividends especially because these theories are consistent with most empirical facts on U.S. firm's payout policy and nobody can deny the work of pioneer researchers, documented in widespread research (Lintner, 1956; Brav et al., 2005; Fama and French, 2001; De Angelo and De Angelo), includes (a) normally mature firms like to pay dividends as compare to young firms; (b) the total payout has been massively grown up through the years; (c) there is a positive correlation between dividends and profits; (d) dividend payout is considered as an important policy in a small number of firms; (e) managers are reluctant to decrease or cut the dividends when firms initiate regular dividends; (f) stock prices go up and down with dividend increase and decrease; (g) dividends changes are asymmetric, with the number of increases exceeding decreases; (h) firms likes to pay dividends through time and they normally do not cumulate excessive cash; (i) it is not necessary that unexpected changes in dividends will predict future surprises in earnings or in profits; and finally, (j) dividend changes as an increase in the dividend refers to high profits and similarly changes as decrease in dividends reflect financial distress and losses.
Another point to be noted is that lifecycle theory is also similar with Modigliani and Miller (1961) views in terms of 'dividends are good'. DeAngelo and DeAngelo (2006) referred the irrelevance proposition of Modigliani and Miller with an argument that their results are restricted by the assumption that free cash flows are 100% distributed each year. According to De Angelo and De Angelo lifecycle theory is a useful contribution in dividend literature in order to solve "dividend puzzle". In upcoming years, it would be attention-grabbing if lifecycle theory resist to new empirical evidence as a new paradigm, in order to replace irrelevance proposition of Modigliani and Miller.
3.4.6 Catering Theory of Dividends
In 2004, Baker and Wurgler developed a new theory named 'Catering Theory of Dividends' by relaxing the core assumptions of Modigliani and Miller regarding perfect and efficient markets. By introducing a simple theoretical model Baker and Wurgler (2004) suggested that (i) a number of investors have uninformed and time varying demand for dividend stock; (ii) managers likes to pay dividends when investors tend to put higher prices on payers, or in other words managers do not pay dividends with investor's preference to non-payers; (iii) arbitrage fails to prevent this demand from driving apart the prices of dividend payers and non-payers.
Baker and Wurgler differentiate the catering theory and clientele effect theory by pointing out that the dividends had not been explored in the past through the investor's sentiments. Clientele effect is related to investor's preferences with respect to desired dividend policy, but in contrast catering theory broadens the level of dividends by predicting the tendency of managers to pay dividends.
As empirical evidence, Baker and Wurgler (2004) introduced the concept of "catering incentives" defined as a measure of market desire for the stocks of dividend payouts. Furthermore, the firm's dividend payout proportion can also be expressed by catering incentives. The team of two authors found the link between catering incentives and dividend payouts. The particular term used by these authors is "dividend premium" defined as a substitute for the value of dividends influenced by the market. This variable is related to dividend payment decision, for instance, if any firm change its dividend policy, there is an expectation of direct impact on investor's sentiments by either about paying or non-paying firms. On the basis of these grounds, Baker and Wurgler (2004) introduced a behavioural model helps to make dividends decisions.
In order to measure the sentiment of relative investor, Baker and Wurgler (2004) found the difference between dividend payers and non payers using logarithm of the book value weighted average market-to-book ratio. In their empirical analysis they also found out that there is a positive correlation between catering incentives and firm's propensity to pay dividends.
Criticism on Catering Theory of Dividends
Cohen and Yagil (2008) try to make an effort to extend the original model of Baker and Wurgler's in their paper "On the catering theory of dividends and the linkage between investment, financing and dividend policies". They explained that their extended model is not only able to predict manager's tendency to pay dividends but it is also eligible to forecast the expected dividend sum as well. Furthermore, the model will be able to find the correlation between investment of the firm, dividend of the firm and financing decisions. According to Cohen and Yagil (2008), three features can be added in the original model.
As a first feature Cohen and Yagil (2008) suggested that relative weight can be measured by the percentage of ownership each type holds. According to them the degree of risk tolerance may not be same for investors, so, relative weight should be the part of the model to maintain the balance. They also explored that Catering Theory of Dividends only deals with manager's tendency to pay dividends, and it does not hold the concept of expected cash dividends. They turned the attention of researcher community towards another important point that Baker and Wurgler assumed that "arbitrageurs" are rational investors with the full knowledge of cost of dividend per share represented as 'C' in the model, but in reality not all the investors have the knowledge of 'C' in technical terms because it carries various values that represent extent of knowledge of cost of dividend that each type of investor has. So, it can be said that 'C' is disaggregated in terms of finding expected dividend per share.
