* Some facts about dividend policy

- Dividends are sticky

- Dividends follow earnings

* Payment Procedures

* Why do firms pay dividends?

- Dividends don't matter: The Miller Modigliani Theorem

- Dividends are taxed heavier than capital gains : Arguments against dividend payments

* Evidence from ex-dividend day price changes

- Dividends are more certain than capital gains: The bird in the hand fallacy

* The Citizen's Utility case

- Dividend policy is tailored to meet clientele needs

* What are clienteles?

* Evidence on clienteles

- Dividends are a good use for excess cash

* Alternatives to dividends

* Arguments for the alternatives

- Dividends as signals

* Need for signalling

* Evidence on dividends as signals

- Dividends as a wealth transfer mechanism

* Rationale

* Evidence

- Management view of dividends

* Framework for analyzing dividend policy


* Dividends are sticky; Firms are much more reluctant to cut dividends than increase them

Measures of Dividend Policy

* Dividend Payout : measures the percentage of earnings that the company pays in dividends

= Dividends / Earnings

* Dividend Yield : measures the return that an investor can make from dividends alone

= Dividends / Stock Price

Three Schools Of Thought On Dividends

1. If

(a) there are no tax disadvantages associated with dividends

(b) companies can issue stock, at no cost, to raise equity, whenever needed

Dividends do not matter, and dividend policy does not affect value.

2. If dividends have a tax disadvantage,

Dividends are bad, and increasing dividends will reduce value

3. If stockholders like dividends, or dividends operate as a signal of future prospects,

Dividends are good, and increasing dividends will increase value

The balanced viewpoint

If a company has excess cash, and few good projects (NPV>0), returning money to stockholders (dividends or stock repurchases) is GOOD.

If a company does not have excess cash, and/or has several good projects (NPV>0), returning money to stockholders (dividends or stock repurchases) is BAD.


* Significant Dates

Declaration date: The dividend is declared at a board of directors meeting. On this date the directors issue a statement similar to the following: " On November 15, 1984, the directors of the XYZ corporation met and declared a regular quarterly dividend of 50 cents per share, plus an extra dividend of 25 cents per share, payable to the holders of record on December 15, payment to be made on January 2, 1985."

Holder-of-record date: At the close of the business on the holder-of-record date, December 15, the company closes its stock transfer books and makes up a list of the shareholders on that date. These shareholders will receive the dividends

Ex-dividend date: Suppose you buy 100 shares on December 13, 1984. Will the company be notified in time? To avoid conflict, the brokerage industry has set up the convention of declaring that the right to the dividend remains with the stock until 4 days prior to the holder-of-record date; on the fourth day before the record date the right to the dividend no longer goes with the stock. This date is called the ex-dividend date. The ex-dividend date in this example in December 11, 1984.

Payment date: The company mails the checks to the recorded holders on January 2, 1985.


The Miller-Modigliani Hypothesis: Dividends do not affect value

Basis: If a firm's investment policy (and hence cash flows) don't change, the value of the firm cannot change with dividend policy. If we ignore personal taxes, investors have to be indifferent to receiving either dividends or capital gains.

* Underlying Assumptions:

(a) There are no tax differences between dividends and capital gains.

(b) If companies pay too much in cash, they can issue new stock, with no flotation costs or signalling consequences, to replance this cash.

(c) If companies pay too little in dividends, they do not use the excess cash for bad projects or acquisitions.

* The Tax Response: Dividends are taxed more than capital gains

Basis: Dividends are taxed more heavily than capital gains. A stockholder will therefore prefer to receive capital gains over dividends.

Evidence: Examining ex-dividend dates should provide us with some evidence on whether dividends are perfect substitutes for capital gains.

Let Pb= Price before the stock goes ex-dividend

Pa=Price after the stock goes ex-dividend

D = Dividends declared on stock

to, tcg = Taxes paid on ordinary income and capital gains respectively

Assume you are all investors in a stock that you bought a long time ago for $P and you have the choice between-

(a) selling before the ex-dividend day, and forsaking the dividend.

(b) selling after the ex-dividend day, and receiving the dividend.

The cash flows from selling before then are-

Pb - (Pb - P) tcg

The cash flows from selling after the ex-dividend day are-

Pa - (Pa - P) tcg + D(1-to)

Since the average investor should be indifferent between selling before the ex-dividend day and selling after the ex-dividend day -

Pb - (Pb - P) tcg = Pa - (Pa - P) tcg + D(1-to)

Moving the variables around, we arrive at the following:

Holding other things equal, the price drop on the ex-dividend day will be equal to the dollar dividend if and only if the marginal investor in the stock faces the same tax rate on dividends and capital gains; it will be less than the dividend if the tax rate on dividends exceeds the tax rate on capital gains; it will be greater than the dividend if the tax rate on dividends is less than the tax rate on capital gains.

If Pb - Pa = D then to = tcg

Pb - Pa < D then to > tcg

Pb - Pa > D then to < tcg

1. Assume that the company that you are analysing has only wealthy individual investors, and that they face a marginal tax rate of 41% on ordinary income, and 28% on capital gains. If the company pays a dividend of $1.00, how much would you expect the price to drop on the ex-dividend day?

What will happen if the capital gains tax rate is lowered to 19.6%, as is being proposed in Congress right now?

The Evidence on Ex-Dividend Day Behavior

The difference between the tax rates on ordinary income and capital gains has changed has changed substantially over time in the United States.

2. Assume that you are a tax exempt investor, and that you know that the price drop on the ex-dividend day is only 90% of the dividend. How would you exploit this differential?

( ) Invest in the stock for the long term

( ) Sell short the day before the ex-dividend day, buy on the ex-dividend day

( ) Buy just before the ex-dividend day, and sell after.

( ) ______________________________________________

Example of dividend capture strategy with tax factors: XYZ company is selling for $50 at close of trading May 3. On May 4, XYZ goes ex-dividend; the dividend amount is $1. The price drop (from past examination of the data) is only 90% of the dividend amount. The transactions needed by a tax-exempt U.S. pension fund for the arbitrage are as follows:

1. Buy 1 million shares of XYZ stock cum-dividend at $50/share.

2. Wait till stock goes ex-dividend; Sell stock for $49.10/share (50 - 1* 0.90)

3. Collect dividend on stock.

Net profit = - 50 million + 49.10 million + 1 million = $0.10 million

Clearly these profits have to exceed transactions costs for this to be worth it. (Transactions costs have to be less than 10 cents per share)

Example of dividend capture strategy even without tax factors

On May 4, 1988 American Electric Power began trading ex-dividend; the dividend amount was $0.565. On May 3, 1988 the following transactions were reported.

10:09:30 am 5,500,000 shares traded at $27.25.

10:09:34 am 2,640,000 shares traded at $26.75

10:09:37 am 2,860,000 shares traded at $26.625

The first transaction represented a buy of 5.5 million shares at $27.25 by a Japanese insurance company (which were then obligated to pay yields of 7-8% to their policy holders from dividend income) from a U.S. pension fund. The second and third transactions represent a sell-back by the same company to the same pension fund of 5.5 million shares at a weighted average price of $26.685 (These were special trades where the pension fund agreed to allow the Japanese firm to collect the dividends of $0.565 on the stock).

Japanese company: was able to collect dividend income of $0.565*5.5 million shares= $3.1 mil

U.S. pension fund: was able to receive the $3.1 million almost 5 weeks early.

* The wrong reasons for paying dividends

A. The bird in the hand fallacy

Argument: Dividends now are more certain than capital gains later. Hence dividends are more valuable than capital gains.

Counter: The appropriate comparison should be between dividends today and price appreciation today. (The stock price drops on the ex-dividend day.)

B. The excess cash hypothesis

Argument: The firm has excess cash on its hands this year, no investment projects this year and wants to give the money back to stockholders.

Counter: So why does not it just repurchase stock? If this is a one-time phenomenon, the firm has to consider future financing needs. Consider the cost of issuing new stock:

 Size of Issue

 Cost of Issue

   Bonds Preferred Stock Common Stock
 Under $ 1 million 14.0% - 22.0%
 $ 1 - $ 1.9 million 11.0% - 16.9%
 $ 2- $ 4.9 million 4.0% - 12.4%
 $ 5 -$9.9 million 2.4% 2.6% 8.1%
 $10 - $ 19.9 million 1.2% 1.8% 6.0%
 $ 20 - $ 49.9 million 1.0% 1.7% 4.6%
 > $ 50 million 0.9% 1.6% 3.5%

* Are firms perverse? Some evidence that they are not

Some investors clearly prefer to receive dividends. Companies with such investors have to pay dividends to keep them happy.

Citizens Utility is a company which has two classes of stock. Class A gets a stock dividend and can be converted freely into Class B stock. Class B gets a cash dividend and cannot be converted to Class A stock. The stock dividend is generally 7% to 13% greater than the cash dividend.

The study found that PB > PA by more than 10%. In other words, the cash dividend shares sold at a premium of 10% over the capital gains shares.

An Updated Study of Canadian companies arrives at similar conclusions.

An updated study of Canadian stocks arrives at similar conclusions; cash dividend shares sell at a premium over stock dividend shares.

Premium on Cash Dividend Shares over

Stock Dividend Shares

Consolidated Bathurst




Dome Petroleum


Imperial Oil


Newfoundland Light & Power


Royal Trustco








3. Clearly some investors like dividends. What types of investors do you think are most likely to fall into this category? (You can pick more than one)

( ) Wealthy investors

( ) Institutional Investors

( ) Less well-off investors

( ) Tax-exempt investors

* A clientele based explanation

Basis: Investors may form clienteles based upon their tax brackets. Investors in high tax brackets may invest in stocks which do not pay dividends and those in low tax brackets may invest in dividend paying stocks.

Evidence: A study of 914 investors' portfolios was carried out to see if their portfolio positions were affected by their tax brackets. The study found that

(a) Older investors were more likely to hold high dividend stocks and

(b) Poorer investors tended to hold high dividend stocks

The following regression captures the determinants of dividend yield

Dividend Yieldt = a + b bt + c Aget + d Incomet + e Differential Tax Ratet + et






Beta Coefficient


Higher beta stocks pay lower dividends.


Firms with older investors pay higher



Firms with wealthier investors pay lower

Differential Tax Rate


If ordinary income is taxed at a higher rate

than capital gains, the firm pays less


4. Assume that you run a phone company, and that you have historically paid large dividends. You are now planning to enter the telecommunications and media markets. Which of the following paths are you most likely to follow?

( ) Courageously announce to your stockholders that you plan to cut dividends and invest in the new markets.

( ) Continue to pay the dividends that you used to, and defer investment in the new markets.

( ) Continue to pay the dividends that you used to, make the investments in the new markets, and issue new stock to cover the shortfall

( ) Other

Can you find a way out of your dilemm

* The Signalling Hypothesis

Context: In a world of asymmetric information, firms have to convince investors about their future prospects. An increase in dividends is one way to signal good future prospects.

On average, at least, increases in dividends seem to be viewed as positive signals and decreases as negative signals.

* Why is the positive reaction to a dividend increase more muted than the negative reaction to a dividend decrease?

An Alternative Story..Dividends as Negative Signals

The problem with signalling stories is that an equally compelling case can be made for increasing dividends being a negative signal, especially for young, high growth firms. There is evidence that earnings growth declines after firms initiate dividends.

Extending the effects of dividend changes on stock prices - long term

While dividend increases (decreases) tend to have a positive (negative) effect in the short term, there seems to be an interesting evening out of returns in the month after the announcement.

* The Wealth Transfer Hypothesis

Basis: Dividends are one way stockholders can transfer wealth from bondholders to themselves. Since bondholders anticipate this, they write constraints into bond covenants and stockholders who do not pay dividends will actually be transferring wealth to bondholders.

CAR: Cumulative Abnormal Return (Actual Return - Expected Return: CAPM)

* Management Beliefs about Dividend Policy
Statement of Management Beliefs


No Opinion


1. A firm's dividend payout ratio affects the price of the stock




2. Dividend payments provide a signaling device of future prospects




3. The market uses dividend announcements as information for

assessing firm value.




4. Investors have different perceptions of the relative riskiness of dividends and retained earnings.




5. Investors are basically indifferent with regard to returns from dividends and capital gains.




6. A stockholder is attracted to firms that have dividend policies appropriate to the stockholders' tax environment.




7. Management should be responsive to shareholders' preferences regarding dividends.





A. Investment Opportunities

Basis: Other thing remaining equal, a firm with more investment opportunities will pay a lower fraction of its earnings as dividends than a stable firm.

Proxy for investment opportunities: Growth rate in firm's assets; Capital Investment;

Testable proposition: A firm with higher growth rates in assets or earnings, and greater capital investment needs will pay out a lower fraction of its earnings as dividends

B. Stability in earnings

Basis: Other things remaining equal, a firm with more stable earnings will pay out a higher fraction of its earnings as dividends than a firm with variable earnings

Proxy for variability in earnings: Variance in EPS

Testable proposition: A firm with higher variance in EPS will have a lower dividend payout ratio

C. Alternative sources of capital

Basis: Other things remaining equal, a firm which can issue new stock or bonds at low cost (such as underwriting commissions) will be more likely to have a high dividend payout ratio.

Proxy for cost of issue: Size of the firm

Testable proposition: A smaller firm will almost invariably have a higher issuance cost than a larger firm in issuing new stock and debt. It will therefore be less likely to have a high payout ratio.

D. Constraints

Basis: Firms which have borrowed large amounts of debt usually have several constraints on their dividend policy and will therefore follow more conservative dividend policies

Proxy for leverage: Debt ratio

Testable proposition: A firm with a high debt ratio will very seldom be able to make make major changes in its dividend policy because of constraints on payout.

E. Signalling Incentives

Basis: Firms which are undervalued may use dividend increases as signals to the market

Proxy for undervaluation: Price/ Value ratios

Testable proposition: As the ratio of price to value decreases dividend increases will become more frequent.

F. Stockholder characteristics

Basis: Firms which have acquired a reputation as high dividend yield firms also acquire stockholders who desire high dividends. Consequently they cannot suddenly shift policy. Testable proposition: The past history of a company's dividend policy is usually be a good indication of what it will do in the future.



Standard Deviation



Free CF to Equity










Dividend Payout Ratio


Cash Paid as % of FCFE




Required Return


ROE - Required return







Standard Deviation



Free CF to Equity










Dividend Payout Ratio


Cash Paid as % of FCFE




Required Return


ROE - Required return







Standard Deviation



Free CF to Equity










Dividend Payout Ratio


Cash Paid as % of FCFE




Required Return


ROE - Required return







Standard Deviation



Free CF to Equity















Dividend Payout Ratio


Cash Paid as % of FCFE




Required Return


ROE - Required return