35
DIVIDEND DECISION
ARIZONA STATE UNIVERSITY
FIN 540 - ADVANCED FINANCIAL MANAGEMENT
WEEK 9
12.1 INTRODUCTION:
Dividends are payments made by a business to its shareholders. In some cases, this
seems to be viewed by directors and shareholders as the equivalent of periodic interest
payments usually made to lenders; compensation for delays in shareholder consumption.
Dividends are also seen as a distribution of a business's current profits to its owners, i.e.
shareholders. A share, like any other economic asset (i.e., an asset whose value is not wholly
or partly derived from sentiment or emotion), is valued based on the future cash flows
expected to arise from the share. Unless a takeover, liquidation, or share buyback is viewed as
a possibility, the only cash flow that may arise from a stock is dividends. So it seems that
dividend anticipation is usually the only determinant of the share price and thus the cost of
equity capital.
Therefore, it seems that directors would be better off increasing shareholder wealth by
paying dividends to the extent allowed by law in certain circumstances. However, this attitude
raises several questions. Is it logical that directors can increase the value of the company's
shares simply by deciding to pay a larger dividend; can value be created in such a simple
way? What if the business has lucrative new investment opportunities; would it be beneficial
to shareholders if the company did not pay dividends in order to have sufficient funds
available to make investments? Alternatively, should the company pay dividends and raise the
necessary funding for the investment opportunity by issuing additional shares to the investor
community, perhaps including existing shareholders who want to get involved? Even when
the company does not see any profitable investment opportunities, should the company be
cautious and pay dividends of less than the full amount allowed by law? What should a
company's dividend policy be and does it matter? We will discuss these questions in this
chapter.
12.2 MODIGLIANI AND MILLER ON DIVIDENDS:
36
In 1961, Modigliani and Miller (MM) published an important article on dividends and
their effect on shareholder wealth (Miller and Modigliani, 1961). Their theme further
developed the Fisher separation principle, which we discussed in Chapter 2. There we saw a
very simple example that suggested that directors need not think about paying dividends.
Provided the company takes on all available investment projects that have a positive NPV,
when discounted at the cost of capital, shareholder wealth will be maximized. The remaining
funds should be paid out to shareholders as dividends.
MM takes it a step further by asserting that the value of a stock will not be affected by
its expected dividend pattern. If shareholders want dividends, they can get them by selling
some of their shareholding. If dividends are paid to shareholders who prefer to leave their
funds in the business, they can cancel the dividends by using the cash received to buy
additional shares in the business, in the capital market. Let's use an example to illustrate MM's
proposition.
Example 12.1 White plc has net assets with a net present value of Rp. 5 billion. This
includes cash of Rp. 1 billion. The directors have identified the possibility that this cash could
be invested in a project whose anticipated inflows have a present value of €2 billion.
Assuming White plc is financed by 1 million ordinary shares (no gearing) and the investment
is made and no dividends are paid, the value of each share should be Rp. (5).
+ If, instead of making an investment, the Rp. 1 Billion in cash was used to pay dividends,
each share would be worth Rp. (5 - 1) Billion/1 Billion
= Rp. 4,000 and the holder of one share will receive Rp. 1,000 (the dividend) in cash.
It will clearly benefit shareholders if the company makes investments, and therefore
investments must be made. If White plc wants to pay dividends (using all available cash) and
make investments, it will need to raise new funds. Assuming that the company wants to
maintain its all-equity status, it will need to issue new shares equal to the value of the
dividend. Since the value of the business after paying dividends and making investments is
Rp. 5 Billion, this would result in a new share issue price of Rp. 5,000 each (200,000 of which
is to meet the Rp. 1 Billion requirement). This means that the initial shareholders would gain
Rp. 1,000 per share through dividends and lose Rp. 1,000 per share, as each share would have
a value of Rp. 5,000 instead of the Rp. 6,000 that would apply if no dividends were paid. In
other words, dividends will have no effect on shareholder wealth. Investing, on the other hand,
makes a significant difference to their wealth.
Creating homemade dividends
Note that, if a new investment is made, the value of the original shares will be Rp.
37
6,000 each. This value is a combination of the share price and the dividend. Regardless of the
amount of dividends paid, individual shareholders can choose how much dividends to take.
Suppose no dividend is paid on White plc shares, but a particular shareholder who owns 100
shares wants cash of, say, Rp. 60,000. The shareholder can create a 'homemade' dividend by
selling 10 shares (at Rp. 6,000 each, their current value).
Canceling dividends
Conversely, assuming that the company pays a dividend of Rp. 1,000 per share (and
issued 200,000 additional new shares at Rp. 5,000 for dividend payment by using the entire
dividend receipt to purchase new shares. Let's consider again a holder of 100 shares who
receives a dividend of IDR 100,000. This cash can be used to buy 20 new shares (at Rp. 5,000
each), increasing the total investment to 120 shares. The value is Rp. 600,000, equal to the
value of 100 shares if no dividend had been paid. Using the entire dividend to purchase new
shares will ensure that our shareholders maintain the same proportion of ownership (120 out of
1.2 million shares) as they had before the dividend and issuance of new shares (100 out of 1.0
million shares).
Dividend as remainder:
Example 12.1 illustrates MM's main point, which is that the key decisions are
investment decisions. The company should make all investments that generate a positive NPV
when discounted at the opportunity cost of shareholder capital. The remaining funds should be
paid out to shareholders, so that they can pursue other opportunities. Thus, dividends are
leftovers. MM argues that it makes no sense for the capital market to value two identical
businesses differently just because of their dividend policy. If a company wants to pay
dividends, it can do so and raise the necessary funding by issuing new equity. Moreover,
individual shareholders who dislike a company's dividend policy can change it to suit their
tastes through 'home-grown' dividend creation or rejection.
However, MM statements rely on several assumptions that are of dubious validity in
the real world. We will discuss the limitations of these assumptions in the next chapter. The
formal derivation of the MM proposition is in the appendix of this chapter.
What MM doesn't say:
Before we proceed to assess what MM has to say about dividends compared to the
traditional view and against the evidence of what actually happens in real life, it might be
38
worth clarifying exactly what they are saying. What they don't say is that it doesn't matter
whether dividends are paid or not, as far as business value is concerned. If a shareholder never
receives dividends or other cash payments in respect of his shares, then the value of those
shares must be zero. What MM is saying is that dividend patterns are irrelevant to valuation. It
doesn't matter when the payment is made, provided:
•
the company invests in all projects that have a positive NPV when discounted at the
shareholder's opportunity cost of capital (and does not invest in projects that have a
negative NPV); and
•
at the end of the day, all investment returns are paid in cash to shareholders.
The 'end' may take a very long time, but in principle this is not a problem. Even if no
cash is expected to be generated until the business is liquidated, this will not cause the value
of the shares to be lower than the ordinary dividend paid out at that time. MM therefore does
not doubt the validity of the dividend model in valuing equity, which we discussed in Chapter
10.
In principle, the timing of a particular dividend does not matter to shareholders,
provided that the next dividend is (1 + r)n times the previous alternative, where r is the
shareholder's expected return (i.e., his or her opportunity cost of capital) and n is the time
interval between the previous dividend and the next dividend.
Suppose a company can pay a dividend now of D. Alternatively, it can invest that
amount now in a project that will generate a single cash inflow after n years, which will be
paid as a dividend at that time. Shareholders will not care whether the dividend is paid now or
in n years, as long as the next dividend is D(1 + r)n. This is because D(1 + r)n has a present
value of D [i.e., D(1 + r)n/(1 + r)n]. So the dividend D now, or D(1 + r)n in n years, should
have the same effect on stock value and shareholder wealth. If the business can invest in
projects that generate inflows greater than D(1 + r)n (i.e., projects with a positive NPV), then
the value of the stock will increase. Of course, as MM points out, an individual shareholder is
always open to obtaining dividends by selling part of his shareholding if the postponement of
dividends does not suit his personal spending plans.
The key point here is that all business funds belong to shareholders. If the company is
able to invest those funds more effectively, on behalf of the shareholders, than the
shareholders themselves can, then it should do so. If it cannot do this, then it should return the
funds to the shareholders and let them make the investment. This is of course the reason why
businesses should use the opportunity cost of shareholder capital as the discount rate in
investment valuation. If the investment potential of the business cannot reach that level, the
39
funds should be passed on to the shareholders, who can reach it.
12.3 TRADITIONAL VIEW OF DIVIDENDS:
Before MM made their assertion, and provided reasonable evidence to that effect, the
view was taken not only that dividends were the primary determinant of share value, but also
that Rp. 1,000 as a cash dividend was worth more than Rp. 1,000 as a cash dividend.
1,000 investments.
Traditionalists will value stocks differently depending on whether or not dividends will
be paid in the near future. Greater value will be given to a stock where dividends are paid
(and, for example, a share issue is made to finance them) than if dividends are not paid. This
view is based on The 'bird in the hand is worth two in the bush' approach, which is the idea
that secure revenue will increase the value of the asset it is sourced from.
Returning to White plc in Example 12.1, if the stock is priced at Rp. 6,000 before the
dividend of Rp. 1,000 per share, then the price of the stock will not fall to Rp. 5,000 ex
dividend, according to traditionalists, but may be, say, Rp. 5,500. Shareholder wealth will thus
increase by Rp. 500 per share as a result of the dividend. The traditional view implies that
dividend payments reduce the capital market's perception of the level of risk attached to future
dividends. This means that the discount rate applied to the expectation of future dividends will
be lower and the market price will increase as a result.
12.4 WHO IS RIGHT ABOUT DIVIDENDS?
The traditional view of dividends seems to suffer from a lack of logic. The only
circumstance where shareholders logically value Rp. 1,000 in cash over Rp. 1,000 in
investments is when they feel that the business is using its retained cash to make unprofitable
investments (i.e., investments that have a negative NPV when discounted at the same rate
based on the opportunity cost of shareholders' capital and taking into account the level of risk
attached to a particular project). Since, at the end of the day, shareholders should receive the
full proceeds of their investment, provided that the company only invests in projects with a
positive NPV, investment is favored over paying dividends.
MMs seem to have logic on their side, yet many businesses seem to behave as if
dividend patterns do matter. Directors seem to devote a fair amount of thought to setting
dividend levels. Most businesses seem to maintain a fairly stable dividend rate, which implies
that they set aside the same amount of dividends each year. However, it would be a
remarkable coincidence if there was the same amount of cash each year but profitable
40
investment opportunities could not be found. We will look at the dividend policies of some
well-known businesses later, in Section 12.6.
MM's assertion that dividend patterns do not matter is, as we have seen, based on
several assumptions. The extent to which these undermine their analysis may explain why the
logic of the MM proposition does not seem to be fully accepted in practice. We will now
review these assumptions.
•
The capital market is frictionless. This implies that there are no transaction costs or
other barriers for investors to behave in practice as expected in theory. Clearly this is
not valid and may be enough to call MM's views into question. MM's proposition
relies on the ability of individual shareholders to sell their shares to generate dividends
or buy shares to eliminate dividends. The fact that transactions on the capital market
entail substantial costs (brokerage fees and so on) means that, for example, receiving a
dividend, on the one hand, and liquidating part of the investment in the stock on the
other, on the other hand, means that the dividend can be eliminated (to create
'homemade' Dividends), on the other hand, are not perfect substitutes for each other.
•
Securities are efficiently priced in the capital market. The evidence (reviewed in
Chapter 9) tends to support the validity of this assumption. This assumption is
necessary, otherwise (in Example 12.1, above), White plc making the investment
would not necessarily increase the share price by €1,000, nor would the dividend
payment decrease the share price by €1,000.
•
Shares can be issued by companies without incurring legal or administrative costs.
This assumption is not valid; it may be important in the current context. MMs see not
paying dividends, on the one hand, and paying dividends and raising the same amount
through share issues, on the other, as equivalent. However, issuing new shares tends to
involve the business in considerable legal and administrative costs (see Chapter 8).
This appears to be an important factor in decision-making regarding the level of
dividend payout.
•
Taxes, both corporate and personal, do not exist. Of course, this is not true. Taxes are
often an important feature. But from a business point of view, a company's tax
obligations do not depend on the level of dividends.
From a shareholder's perspective, the main question is whether it is more tax efficient to
receive dividends or capital gains. The answer tends to depend on who the shareholder is and
the level of income and capital gains respectively in the year in question. For some
shareholders (such as tax-exempt institutions like pension funds), neither dividends nor capital
41
gains will attract tax. For others, tax is important. To the extent that it is possible to
generalize, it is probably correct to say that for individuals with high levels of dividends and
capital gains, further dividends are likely to be taxed at the same rate as capital gains.
One feature of the UK investment scene is the increasing trend towards share
ownership by investment institutions. Many of these, such as pension and life assurance
funds, are exempt from tax on both income and capital gains.
The situation is further complicated by the fact that a company's tax liability will, to a
large extent, depend on what it does with the funds it saves. Some investments in real assets
are more tax efficient than others; for example, investments in machinery will get tax breaks,
whereas investments in office blocks will not.
The 'no tax' MM assumption may not be very significant provided that companies
show consistency in their dividend policy. We will come back to this point shortly.
The weakness of the individual assumption does not seem to be large enough to destroy the
case for MM. Clearly, the presence of significant transaction and share issuance costs should
weaken it, but there seems to be no reason why MM's analysis should not adequately
represent the true position - at least, as far as their analysis goes.
12.5 OTHER FACTORS:
There are several other factors that may influence this position and explain why
directors seem to consider dividend decisions as important, despite MM's assertion that they
are not.
Dividend information content:
Some take the view that dividend levels, and, perhaps more specifically, changes in
dividend levels, convey new public information about the business. For example, an increase
in the dividend amount may be (and apparently is often interpreted as) a signal that the
directors view the future cash inflows of the company with confidence. This is in stark
contrast to the conclusion that might be reached by following MM analysis. Paying any
dividend, let alone one that shows an increase over the previous year, might indicate that the
directors were unable to find enough investment opportunities to use all the finance available
to them, so they returned some of it to shareholders. This does not indicate great confidence in
the future on the part of the directors. Assuming that an increase in dividends is seen as a
positive sign, whether the signal is intended or unintended, may vary from case to case. Of
course, a common tactic of the management of a takeover-reluctant target business is to
42
increase its dividend rate; perhaps this is a deliberate signal to inspire shareholder confidence
in the future of the business.
If the dividend increase is intended as a signal, it is reasonable to ask why the directors
did not simply issue a statement. Surely a statement would be less ambiguous than a dividend
increase. However, it seems that actions are more meaningful than words (see, for example,
Vieira and Archbold, 2008). Incidentally, if the signaling view on dividends is correct, then an
increase in dividends is expected to have a favorable impact on stock prices in the capital
market. Such a phenomenon would be further evidence that capital markets are not efficient in
the strong form (see Chapter 9), as it implies that information available to directors cannot be
captured in stock prices.
Client effect:
It is widely believed that investors have a preferred habitat, i.e. the type of investment
that they feel suits them best. In the context of dividends, their preferences may be determined
by their respective tax positions. Someone with a high marginal income tax rate would
probably be attracted to stocks with low dividend payouts. On the other hand, investors such
as pension funds that are tax-exempt, yet require regular cash receipts in order to meet
payments to retirees, may prefer to hold shares in businesses with relatively high dividend
payments. Such investors can of course make money by selling shares, but this will involve
brokerage fees and the like, which is best avoided.
If there is a client effect, this means that some of the shareholders
A business acquires its shares because they match the dividend policy of the business. If there
is inconsistency in the dividend policy, then many investors will delay their share purchase
because they do not know whether the dividend amount matches their preference or not. The
lack of popularity of the stock will have an adverse effect on its price and, therefore, on the
cost of capital.
Even if a particular business is fairly consistent in its dividend policy, but then makes a
major change, shareholders who were very fond of the previous dividend policy (this is
probably all of them) will probably seek to move to the shares of the business with a dividend
policy that is more acceptable to them. While new clients may find the dividend policy
attractive, the friction caused by a group of investors selling their shares to a new group of
investors will adversely affect shareholders. Not only that, the uncertainty in the minds of
investors that such a change might happen, regarding how consistent the dividend policy will
be in the future, can also have an impact on the share price.
43
Liquidity:
There is an opinion that the level of dividends paid by a given business, at a given
time, is largely determined by the amount of cash available. This is of course what MM
suggests. On the other hand, if failure to pay dividends is interpreted negatively by the capital
markets, then the best interests of shareholder wealth may be served by ensuring that cash is
available, perhaps by borrowing, or even by passing up profitable investment opportunities.
Target capital gearing and trade-off theory:
As we saw in Chapter 11, there is clear evidence that most businesses tend to have a
targeted capital gearing ratio. Individual firms set this target by trying to balance the benefits
of borrowing (essentially, the tax deductibility of borrowed interest), on the one hand, and the
potential downside of 'bankruptcy', on the other. The existence of a target has implications for
dividend policy where dividend payments reduce equity. In certain cases, this may move the
business away from its target and cause the business to be burdened with debt financing. On
the other hand, paying out dividends can bring another business back to its target. The
alternative theory, the pecking order theory, is not really supported by evidence, as we saw in
Chapter 11.
Agent
The theoretical principle of dividend decisions is clear, according to MM. If the
company cannot identify a profitable investment (positive NPV), then it should return the
remaining cash to shareholders and let them use it more profitably than they can use it in their
favor. However, in an effort to expand its empire, the directors may choose to make an
adverse investment. As we discussed in Chapter 2, the very many senior managers are granted
stock options. This means that these managers have a very clear interest in the potential
market price of the stock, but no direct financial interest in dividends. But for shareholders, a
certain combination of dividend income and share price may be beneficial. There is therefore
a danger that the agency problem of separating the ownership of the business from the
directors may lead to agency costs being borne by the owners - the shareholders.
In short, agency problems can lead to smaller dividends than shareholders would
otherwise receive. As with agency problems, shareholders often do not have the necessary
information that could lead them to raise a reasoned objection to the directors' dividend
decision.
Share repurchase
44
One way for a business to transfer funds to shareholders is by buying back shares from
individual shareholders, either through the stock exchange or by making direct contact with
them. We should be clear that, in essence, this has the same overall impact on the business as
a dividend payment which involves the same amount of cash. The impact on shareholders is
very different and this is the main attraction. Shareholders who wish to withdraw some or all
of their stake in the business (sell some or all of their shares to the business) can do so to the
extent they wish; those who do not wish to sell need not do so. Share buybacks allow
companies to make large cash distributions in any given year without setting any expectations
regarding future dividend levels.
It is common for businesses to conduct share buybacks
that involves a limited number of shareholders while paying a 'normal' dividend to all
shareholders. As we will see later in this chapter, share buybacks have grown in popularity in
recent decades.
12.6 DIVIDEN:
Importance of dividend decision
It appears that managers see the decisions they make regarding the amount and timing
of dividend payments, while important to them, as ranking behind decisions relating to new
investment and business financing. Cohen and Yagil (2010) surveyed a number of managers
of large businesses across different industries in selected countries. Managers were asked to
rate the importance of various decision areas on a five-point scale (1 = not important to 5 =
very important). The mean (average) score for dividend decisions was 2.73, placing dividend
decisions roughly midway between 'not important' and 'very important'. This compares to
scores of 4.13 for investment decisions and 3.81 for financing decisions.
Dividend scale
Whatever the impact of a particular dividend policy on overall shareholder wealth, the
evidence suggests that historically, in the US, dividends have been accounted for 52 percent
of shareholder returns over the period 1872 to 2000, the remaining 48 percent coming from
increases in share value (Coggan, 2001). These proportions have not been consistent over the
period. In recent years, for example, the contribution of dividends has been somewhat smaller
than the average over the period. Between 1995 and 2000, the contribution of dividends was
only 20 percent. There is no reason to believe that the relationship between capital gains and
dividends is significantly different in the UK.
Benito and Young (2003) examined dividends paid by UK companies between 1974
45
and 1999. They found that, since 1995, there has been an increasing trend for firms not to pay
dividends. In 1995, 14.3 percent of companies did not pay dividends, but this number
increased to 25.2 percent in 1999. Although the single most important factor in failing to pay
dividends is lack of profitability, the recent increase in the proportion of firms failing to pay
dividends is attributed to new, growing businesses that have never paid dividends. This last
point is entirely consistent with MM: the finance required for profitable investments is not
used to pay dividends. Benito and Young also note an increasing trend of dividend reduction
over the past few years. On the other hand, higher dividend payers in 1999 paid twice as much
in dividends, compared to sales revenue, as their high dividend-paying counterparts in 1977.
This finding suggests a much more flexible dividend policy in recent years; more non-payers
and larger large payouts. This would be consistent with MM.
Fama and French (2001) found that, in the US, the percentage of companies paying
dividends fell from 66.5 percent in 1978 to 20.8 percent in 1999. Like Benito and Young in
the UK, they found that this decline was partly related to the changing profile of stock market
listed companies to younger, smaller companies with low profitability and strong growth
opportunities. They also found a general decline in dividends across all businesses.
DeAngelo, DeAngelo and Skinner (2004), again with US data, support the findings of
Fama and French that there has been a decline in the number of businesses paying dividends.
However, they also found that the total dividends paid by all businesses have increased. The
larger dividends paid by large payers tend to outnumber other dividends, where there are
lower dividends or no dividends at all. These two US studies are consistent with Benito and
Young's findings in the UK that there seems to be more flexibility in dividend policy than
ever before.
This point of flexibility is further supported by Julio and Ikenberry (2004). They found
that the tendency of American businesses to pay dividends in the late 20th century reversed in
the early 2000s. They attribute this reversal to several factors, among others:
•
the tendency for younger and smaller businesses (mentioned by Benito and Young and
by Fama and French (see above)) to now become larger and more profitable with less
investment capital requirements, leaving them with cash to spare;
•
businesses using dividends as a means to convince investors that their profits are
genuine, following corporate governance scandals such as Enron (discussed in Chapter
3); and
•
more favorable tax treatment of dividends in the US.
•
Cohen and Yagil (2010) found that a higher proportion of businesses pay dividends if:
46
•
the activity is considered less risky (higher debt credit rating); and
•
each business is bigger.
Another factor leading to lower dividends seems to be the agency issues associated with
managers' stock options, which we discussed in Section 12.5. Various researchers, including
Chetty and Saez (2005) and Shapiro and Zhuang (2013), find a tendency for businesses whose
managers hold large amounts of stock options to pay lower dividends.
Dividend policy stability
Evidence from casual observation is that directors seek to maintain dividends at
previous levels; dividend reductions appear to be a relatively rare occurrence.
This is strongly supported by more formal evidence. Vieira and Archbold (2008)
conducted a survey (in 2003) of UK listed companies. They find that managers feel:
•
future income is an important determinant of dividend levels (83 percent of
respondents);
•
historical dividend patterns are important or very important (73 percent);
•
dividend cuts should be avoided (87 percent); And
•
dividend rate changes are important or very important (79 percent).
Baker, Powell and Veit (2002) conducted a similar survey of managers of large American
businesses. They found similar results. Cohen and Yagil (2010) found that on average 36
percent of businesses (across industries and countries) pay a constant dividend per share, and
it will probably increase over time. About 25 percent of companies pay a fixed percentage of
their net income, 33 percent pay no dividends, and only 6 percent follow other dividend
policies.
The effect of dividends on stock prices
Early studies on the relationship between stock prices and dividend policy (such as
Friend and Puckett, 1964; and Black and Scholes, 1974) failed to find evidence that higher or
lower dividend rates lead to higher or lower returns either before or after taxes. However,
Fama and French (1998), based on the US experience, showed such a relationship. These
researchers concluded that dividend increases are treated by the market as a signal of higher
expectations from managers. This tends to lead to an increase in share prices as investors
reassess the value of the business. Fama and French's findings are supported by other
researchers, including Akbar and Stark (2003), who based their study on UK businesses, and
Hand and Landsman (2005), using US data. While there seems to be doubt regarding the
impact of dividend increases on stock prices, dividend decreases always seem to be bad news
47
for stock prices (see for example, Healy and Palepu, 1988).
Dividend information content (signal):
Pettit (1972) found clear support for the proposition that the capital market takes into
account dividend announcements as information to assess stock prices. This finding is quite
consistently supported by subsequent studies, for example Aharony and Swary (1980). Nissim
and Ziv (2001) find that dividend changes provide information about the level of profitability
in the years following the change. Dividend increases tend to show an increase in accounting
profit in the two years following the dividend increase. This is related to the findings of Fama
and French (1998). The evidence on the informational impact of dividends seems to be quite
convincing. Mougoue and Rao (2003) find that not all businesses signal future profitability
through dividend policy; only about a quarter of the businesses they examine do so. They note
that 'signalers' tend to be smaller, have lower asset growth rates and have higher than average
capital gearing.
Dewenter and Warther (1998) compared the signaling effects in America and Japan.
They conclude that dividends seem to have greater information content in America than in
Japan. They conclude that this arises from the closer relationship between investors and
managers in Japan, which leads to better information flows from investors and managers in
Japan. This may indicate reduced agency problems in Japan. Lack of information by
shareholders tends to be the main cause of agency costs.
Hand and Landsman (2005) (H and L) contradict the common view on signaling. They
analyzed the dividend/stock price relationship for a large number of US businesses between
1984 and 1996. They concluded that businesses do not deliberately use dividends to signal,
but that the size of dividends allows investors to reassess their previous projections of future
earnings and cash flows. H and L imply that investors tend to underestimate these projections
and that dividends often reflect this, leading to a corresponding rise in share prices. If H and L
are correct, then it is not difficult to see why previous researchers misinterpreted the role of
dividends in this context. This also solves the puzzle of why managers don't simply tell
investors how their business is doing, instead sending signals that may be difficult to interpret;
according to H and L they are does not send such signals.
Client effect:
Elton and Gruber (1970) conducted quite an interesting study to test the client effect.
By looking at the fall in share price when a stock goes from cum dividend to ex dividend
48
(shortly before the dividend payment), they were able to deduce the average marginal income
tax rate for shareholders of a particular business. They found that lower income tax rates were
associated with stocks paying high dividends and higher income tax rates with stocks paying
low dividends: that is, they found a clientele effect.
Pettit (1977) gained access to information on the securities portfolios of a large
number of clients of a major US stockbroker. He found that low dividend rates seem to be
preferred by investors with relatively high marginal income tax rates, by young investors, and
by those who are less risk averse on average.
However, Lewellen, Stanley, Lease and Schlarbaum (1978), using the same data but a
different approach from that taken by Pettit, reached a different conclusion that there is only a
very weak client effect. Litzenberger and Ramaswamy (1982) conducted a study that
produced results that seemed to support the client effect. Crossland, Dempsey and Moizer
(1991), using UK data, also found clear evidence of a significant client effect. Graham and
Kumar (2006) looked at the investment behavior of 60,000 American households and found a
clear client effect.
Allen, Bernardo and Welch (2000) identified an interesting relationship between
client effects and agency problems. They argue that businesses that pay dividends tend to
attract institutional investors, many of whom are tax-exempt (life insurance funds, pension
funds, and the like). Institutions, as they tend to be large and professionally managed
investors, tend to take steps to ensure that the businesses they invest in are well managed.
Various studies on the client effect broadly seem to support its existence.
Corporate tax:
A study conducted by Siddiqi (1995) shows, as expected, that changes in tax laws that
make dividends cheaper for businesses tend to be associated with an increase in dividends.
This is supported by Pattenden and Twite (2008). This is also supported by Jacob and Jacob
(2013) who examined dividend payments by 6,035 businesses in 25 countries over the period
1990 to 2008. They found that different tax treatments of conventional dividends and share
buybacks have a large impact on business distributions.
Cash flow and investment opportunities:
Barclay, Smith and Watts (1995) found, based on a sample of businesses in the US,
that dividends tend to be lower when there are more investment opportunities for the business,
as suggested by MM principles. Cohen and Yagil (2010) found that expected future cash
49
flows are the most important factor for managers when deciding their dividend policy.
Vieira and Archbold (2008) found that 41 percent of the UK managers they surveyed
rated the availability of good investment opportunities as an important, or very important,
influence on dividend decisions. Despite this, 63 percent agreed, or strongly agreed, that their
business would raise new funds to support investment opportunities, rather than reduce
dividends as a means to finance such opportunities. Clearly, businesses tend to be very
unwilling to reduce dividend levels, once set.
Examples of real businesses that follow MM principles:
Online fashion and beauty retailer ASOS plc was formed in 2004. The business has yet
to pay a dividend, although it is still making a profit every year up to the time of writing. In its
2015 annual report, the company said: One of the challenges is making the right investments
to build the necessary infrastructure to achieve our next goals. We accelerated our investment
plans last year and will do so again this year. Our next staging post is Rp. 2.5 trillion in sales
and we are reinvesting our money to set up the logistics and technology infrastructure to meet
this. Board again decided not to declare a dividend.
Michael O'Leary, the CEO of the colorful airline Ryanair Holdings plc, is more direct
than ASOS plc. As reported by Osborne (2004), he said: We will never pay dividends as long
as I live and breathe and as long as I am the largest shareholder. If you are foolish enough to
invest in an airline for a dividend stream, you should be put back where you came from.
More diplomatically, the business' website states: Ryanair Holdings anticipates, for the
foreseeable future, that it will retain future earnings to fund the company's business
operations, including the acquisition of additional aircraft required for Ryanair's planned entry
into new markets and expansion of existing services, as well as replacement aircraft for its
current fleet. Ryanair Holdings anticipates that it will retain future earnings to fund its
business operations, including the acquisition of additional aircraft required for Ryanair's
planned entry into new markets and expansion of existing services, as well as replacement
aircraft for its current fleet.
In fact, Ryanair bought back some shares in 2008 and paid a dividend (the first one) in
September 2010. This followed the company's decision not to expand its fleet as planned,
probably because the company did not foresee any further investments in the near future. -
ment has a positive net present value. This left the business with a large amount of uninvested
cash. The company has repurchased shares and/or paid dividends almost every year since
2008. However, the business is adamant that shareholders should not expect to receive a
50
distribution in any year. This depends on cash resources and investment demand. These are
two examples of businesses that seem to follow the MM approach. It must be said that such
businesses seem to be in the minority. Most businesses do not see dividends as a residual
when investment needs have been exhausted.
Examples of real businesses that don't follow MM principles:
Most businesses seem to have a dividend policy that is not linked to investment
opportunities. For example, The Go-Ahead Group plc (which is in the transportation business
- owning many local bus and rail operations), according to its 2015 annual report, has a policy
of paying half of its after-tax profits as dividends. Greene King plc, an operator of breweries,
pubs and hotels, also has a target of paying dividends of twice its profits (2015 annual report).
Perhaps surprisingly, Ettridge and Kim (1994) find that when changes in accounting
methods lead to higher reported earnings, as well as higher actual tax expenses, then dividend
levels seem to increase. This seems strange, as investment opportunities and the level of
operating cash flows are not affected by accounting changes. Higher tax payments are the
only real impact of the change. This implies that businesses link dividends more to reported
accounting profits than to the underlying economic reality.
Share buyback and MM:
Share buybacks became a very common activity in the UK, probably until the
recession hit in 2008. Oswald and Young (2002) looked at buybacks by UK listed companies
over the period 1995 to 2000. They found that repurchase activity in 2000 was seven times
that in 1995. After 2008, buybacks declined rapidly such that buybacks in the UK in 2009
represented only 3 percent of the level of buybacks in 2008 (Secker, 2010). Recently,
however, repurchase activity appears to be increasing. Benhamouda and Watson (2010)
investigated the motivations for share repurchases among 267 businesses that had made
repurchases in the early 2000s. They found that repurchases tended to be made:
•
By big companies;
•
By companies with relatively high dividend payout rates; and
•
If there are large profit figures, especially 'windfall' profits
unexpected.
This finding was interpreted by the researchers as an indication that businesses were opting
for buybacks in lieu of dividends. The sharp decline in the popularity of buybacks towards the
end of the 2000s seems to be linked to the sharp decline in corporate profitability, which was
51
also linked to the recession.
In the US, share buybacks are also increasingly popular, at the expense of conventional
dividends (Koller, 2015). Since the early 1990s, large US companies have consistently
distributed around 85 percent of their profits to shareholders, but the balance between
dividends and share buybacks has changed dramatically. Whereas in the early 1980s only 10
percent of profits were distributed as share buybacks, this figure has increased to 23 percent in
the early 1990s and to 47 percent in the early 2010s. Conventional dividends have contracted
commensurately.
On the face of it, a share buyback is an example of a business following MM
principles and returning shareholders funds that they cannot profitably invest on behalf of
those shareholders.
Nissan Motor Corporation bought back shares worth Yen 400 billion during 2016.
This amount represents about 7 percent of its issued shares. The business had expanded most
of its production capacity and had cash left over, while at the same time, its share price was
slumping and Nissan wanted to improve this. Logically, reducing the number of shares issued
would increase the share price as future earnings would be associated with fewer shares. In
returning the funds to shareholders, Nissan must have thought that those investors could
invest, outside the business, at a higher rate of return than Nissan could achieve for them
inside the business.
Upscale and online fashion and home fabrics business Next plc bought about Rp.
140 billion of its shares in early 2016. The aim seems to have been primarily to support the
slumping share price following disappointing trading results over Christmas 2015.
Agent:
Agency problems (and costs) tend to arise when shareholders and directors have
different objectives. Schooley and Barney (1994) provide some evidence that when senior
managers are also shareholders, agency problems, in the context of dividends, tend to be
reduced. See also the work of Allen, Bernardo and Welch (2000), mentioned in the client
effects section above.
Dividend reinvestment:
MM states that shareholders can negate the impact of dividends by reinvesting them in
the business. Some UK businesses make this relatively easy for shareholders to do. For
example, BP plc (oil), Rolls-Royce plc (engine manufacturing), and Kingfisher plc (DIY
52
retailer) give their shareholders the option to choose dividends in cash or additional shares.
A number of companies will use dividend cash to buy their own shares, on the Stock
Exchange, on behalf of shareholders who want this done for them, at a low cost to
shareholders. These include, for example, Royal Dutch Shell plc (oil), Unilever plc
(household products), Rio Tinto plc (mining) and TUI Travel plc (holidays).
12.7 CONCLUSION ON DIVIDENDS:
Regarding capital gearing, we do not know whether the traditionalists' or MM's
assertions are correct. However, we have some empirical evidence on this matter so we may
be able to reconcile the two views.
MM's assertion that dividend patterns are irrelevant has some credibility. The
assumption is not entirely convincing, but it does not seem plausible enough to completely
invalidate the irrelevance assertion. However, the evidence does not seem to supporting MM's
view, in practice the valuation seems to be linked to the dividend pattern.
It is possible that MM is generally correct when he says that dividend patterns are not
affect value, unless there is evidence that dividends change investors' perception of the future
of the business.
From the available evidence, there appears to be a clientele effect whereby dividend
policies linked to specific stocks appear to attract an identifiable group of investors.
Recognition of the clientele effect by directors could explain their reluctance to change
dividend payout levels. Given that the client effect appears to be significant, investors need to
be aware of a business's dividend policy. Perhaps the most effective way to inform investors
of this policy is to establish a fairly constant pattern and stick to it. Failure to demonstrate
consistency is likely to lead to uncertainty for investors and the need for shareholders to
abandon one habitat in favor of another, more favorable one. In either case, a change in
dividend policy is likely to lead to lower share prices and a higher cost of capital. All this
seems to imply that liquidity may not be an important factor in dividend decisions. Perhaps,
businesses will ensure that there is enough cash available to pay out dividends, one way or
another.
Traditional view:
•
Dividend decision - an important decision for the impact of shareholder wealth.
•
Shareholders value dividends more highly than the equivalent amount retained in the
business.
Modigliani and Miller (MM) view.
53
•
Dividends should be paid only if the company cannot use the available funds at least
as effectively as the shareholders can, i.e. invest in all projects with a positive NPV
when discounted by the opportunity cost of shareholders' capital.
•
Only the funds remaining after investment should be paid out as dividends - dividends
are residual.
•
The share price is the PV of future dividends - shareholders will be indifferent to how
the PV is created (the size of the dividend each year). If paying less this year will yield
more in subsequent years so that the PV (and share price) increases, then shareholder
wealth will increase.
•
Individual shareholders can customize the business's dividend payout to meet their
needs.
•
If there are no (or few) dividends and they need funds, they can sell some of their
shares to generate cash inflows from them (a 'homemade' dividend).
•
If they choose not to receive dividends from the business, they can buy more shares
with the cash from the dividends and, thus, leave their investment intact.
