## CHAPTER 16

The Dividend Controversy

1.                  Newspaper exercise; answers will vary depending on the stocks chosen.

2.                  The available evidence is consistent with the observation that managers believe shareholders prefer a steady progression of dividends.  For managers of risky companies whose earnings have high variability, it is easy to show, using the Lintner model, that a lower target payout (e.g., zero) and a lower adjustment rate (e.g., zero) reduce the variance of dividend changes.

3.         a.         Distributes a relatively low proportion of current earnings to offset fluctuations in operational cash flow; lower P/E ratio.

b.                  Distributes a relatively high proportion of current earnings since the decline is unexpected; higher P/E ratio.

c.                  Distributes a relatively low proportion of current earnings in order to offset anticipated declines in earnings; lower P/E ratio.

d.         Distributes a relatively low proportion of current earnings in order to fund expected growth; higher P/E ratio.

4.         a.         A t = 0 each share is worth \$20.  This value is based on the expected stream of dividends: \$1 at t = 1, and increasing by 5% in each subsequent year.  Thus, we can find the appropriate discount rate for this company as follows:

Þ        r = 0.10 = 10.0%

Beginning at t = 2, each share in the company will enjoy a perpetual stream of growing dividends: \$1.05 at t = 2, and increasing by 5% in each subsequent year.  Thus, the total value of the shares at t = 1 (after the t = 1 dividend is paid and after N new shares have been issued) is given by:

If P1 is the price per share at t = 1, then:

V1 = P1 ´ (1,000,000 + N) = \$21,000,000

and:

P1 ´ N = \$1,000,000

From the first equation:

(1,000,000 ´ P1) + (N ´ P1) = 21,000,000

Substituting from the second equation:

(1,000,000 ´ P1) + 1,000,000 = 21,000,000

so that P1 = \$20.00

b.                  With P1 equal to \$20, and \$1,000,000 to raise, the firm will sell 50,000 new shares.

c.                  The expected dividends paid at t = 2 are \$1,050,000, increasing by 5% in each subsequent year.  With 1,050,000 shares outstanding, dividends per share are: \$1 at t = 2, increasing by 5% in each subsequent year.  Thus, total dividends paid to old shareholders are: \$1,000,000 at t = 2, increasing by 5% in each subsequent year.

d.                  For the current shareholders:

5.                  From Question 4, the fair issue price is \$20 per share.  If these shares are instead issued at \$10 per share, then the new shareholders are getting a bargain, i.e., the new shareholders win and the old shareholders lose.

As pointed out in the text, any increase in cash dividend must be offset by a stock issue if the firm’s investment and borrowing policies are to be held constant.  If this stock issue cannot be made at a fair price, then shareholders are clearly not indifferent to dividend policy.

6.                  The risk stems from the decision to not invest, and it is not a result of the form of financing.  If an investor consumes the dividend instead of re-investing the dividend in the company’s stock, she is also ‘selling’ a part of her stake in the company.  In this scenario, she will suffer an equal opportunity loss if the stock price subsequently rises sharply.

7.                  No, this does not make sense.  Restricting dividends does not restrict the investor’s ‘wages.’  For the policy to be effective, it would also have to restrict capital gains.

8.                  If the company does not pay a dividend:

 Cash 0 0 Debt Existing fixed assets 4,500 5,500 + NPV Equity New project 1,000 + NPV \$5,500 + NPV \$5,500 + NPV

If the company pays a \$1,000 dividend:

 Cash 0 0 Debt Existing fixed assets 4,500 1,000 Value of new stock New project 1,000 + NPV 4,500 + NPV Value of original stock \$5,500 + NPV \$5,500 + NPV

Because the new stockholders receive stock worth \$1,000, the value of the original stock declines by \$1,000, which exactly offsets the dividends.

9.                  One problem with this analysis is that it assumes the company’s net profit remains constant even though the asset base of the company shrinks by 20%.  That is, in order to raise the cash necessary to repurchase the shares, the company must sell assets.  If the assets sold are representative of the company as a whole, we would expect net profit to decrease by 20% so that earnings per share and the P/E ratio remain the same.  After the repurchase, the company will look like this next year:

 Net profit: \$8 million Number of shares: 0.8 million Earnings per share: \$10 Price-earnings ratio: 20 Share price: \$200

10.       a.         If we ignore taxes and there is no information conveyed by the repurchase when the repurchase program is announced, then share price will remain at \$80.

b.                  The regular dividend has been \$4 per share, and so the company has \$400,000 cash on hand.  Since the share price is \$80, the company will repurchase 5,000 shares.

c.                  Total asset value (before each dividend payment or stock repurchase) remains at \$8,000,000.  These assets earn \$400,000 per year, under either policy.

Old Policy:  The annual dividend is \$4, which never changes, so the stock price (immediately prior to the dividend payment) will be \$80 in all years.

New Policy:  Every year, \$400,000 is available for share repurchase.  As noted above, 5,000 shares will be repurchased at t = 0.  At t = 1, immediately prior to the repurchase, there will be 95,000 shares outstanding.  These shares will be worth \$8,000,000, or \$84.21 per share.  With \$400,000 available to repurchase shares, the total number of shares repurchased will be 4,750.  Using this reasoning, we can generate the following table:

 Time Shares Outstanding Share Price Shares Repurchased t = 0 100,000 \$80.00 5,000 t = 1 95,000 \$84.21 4,750 t = 2 90,250 \$88.64 4,513 t = 3 85,737 \$93.31 4,287

Note that the stock price is increasing by 5.26% each year.  This is consistent with the rate of return to the shareholders under the old policy, whereby every year assets worth \$7,600,000 (the asset value immediately after the dividend) earn \$400,000, or a return of 5.26%.

11.       If markets are efficient, then a share repurchase is a zero-NPV investment.  Suppose that the trade-off is between an investment in real assets or a share repurchase.  Obviously, the shareholders would prefer a share repurchase to a negative-NPV project.  The quoted statement seems to imply that firms have only negative-NPV projects available.

Another possible interpretation is that managers have inside information indicating that the firm’s stock price is too low.  In this case, share repurchase is detrimental to those stockholders who sell and beneficial to those who do not.  It is difficult to see how this could be beneficial to the firm, however.

12.             Because companies are reluctant to reduce their dividends, they will normally increase dividends only when management is fairly certain that the increases can be sustained.  Thus, an increase in dividends signals management’s confidence about the company’s future earnings potential, and it is this signal that causes the stock price to rise.

13.       a.         This statement implicitly equates the cost of equity capital with the stock’s dividend yield.  If this were true, companies that pay no dividend would have a zero cost of equity capital, which is clearly not correct.

b.                  One way to think of retained earnings is that, from an economic standpoint, the company earns money on behalf of the shareholders, who then immediately re-invest the earnings in the company.  Thus, retained earnings do not represent free capital.  Retained earnings carry the full cost of equity capital (although issue costs associated with raising new equity capital are avoided).

c.                  If the tax on capital gains is less than that on dividends, the conclusion of this statement is correct; i.e., a stock repurchase is always preferred over dividends.  This conclusion, however, is strictly because of taxes.  Earnings per share is irrelevant.

14.       a.         If we assume that the constraint on dividends is binding, that is, if we assume that dividends would have risen in the absence of the constraint, then the restriction on dividends would increase capital gains.  In other words, the total return to shareholders would not change.  Dividends would be lower than otherwise, but capital gains would increase to offset the reduction in dividends.  Thus, stock prices would increase.

b.                  The total return to equity capital is unchanged, and the firm’s overall cost of capital is also unchanged.  Thus, there will be no effect on capital investment.

15.       a.         Because this is a regular dividend, the announcement is not news to the stock market.  Hence, the stock price will adjust only when the stock begins to trade without the dividend and, thus, the stock price will fall on the ex-dividend date.

b.                  With no taxes, the stock price will fall by the amount of the dividend, here \$1.

c.                  With taxes on dividends but no taxes on capital gains, investors will require the same after-tax return from two comparable companies, one of which pays a dividend, the other, a capital gain of the same magnitude.  The stock price will thus fall by the amount of the after-tax dividend, here \$1 ´ (1 - 0.30) = \$0.70.

d.                  If dealers are taxed equally on capital gains and dividends, then they should not demand any extra return for holding stocks that pay dividends.  Thus, if shareholders are able to freely trade securities around the time of the dividend payment, there should be no tax effects associated with dividends.

16.       a.         If you own 100 shares at \$100 per share, then your wealth is \$10,000.  After the dividend payment, each share will be worth \$99 and your total wealth will be the same: 100 shares at \$99 per share plus \$100 in dividends, or \$10,000.

b.                  With no taxes, it does not matter how the company transfers wealth to the shareholders; that is, you are indifferent between a dividend and a share repurchase program.  In either case, your total wealth will remain at \$10,000.

17.             After-tax Return on Share A:  At t = 1, a shareholder in company A will receive a dividend of \$10, which is subject to taxes of 30%.  Therefore, the after-tax gain is \$7.  Since the initial investment is \$100, the after-tax rate of return is 7%.

After-tax Return on Share B:  If an investor sells share B after 2 years, the price will be: (100 ´ 1.102) = \$121.  The capital gain of \$21 is taxed at the 30% rate, and so the after-tax gain is \$14.70.  On an initial investment of \$100, over a 2-year time period, this is an after-tax annual rate of return of 7.10%.

If an investor sells share B after 10 years, the price will be:

(100 ´ 1.1010) = \$259.37.  The capital gain of \$159.37 is taxed at the 30% rate, and so the after-tax gain is \$111.56.  On an initial investment of \$100, over a 10-year time period, this is an after-tax annual rate of return of 7.78%.

18.       a.         (i)   The tax-free investor should buy on the with-dividend date because the dividend is worth \$1 and the price decrease is only \$0.90.

(ii)   The dividend is worth only \$0.60 to the taxable investor who is subject to a 40% marginal tax rate.  Therefore, this investor should buy on the ex-dividend date.  [Actually, the taxable investor’s problem is a little more complicated.  By buying at the ex-dividend price, this investor increases the capital gain that is eventually reported upon the sale of the asset.  At most, however, this will cost:

(0.16 ´ 0.90) = \$0.14

This is not enough to offset the tax on the dividend.]

b.                  The marginal investor, by definition, must be indifferent between buying with-dividend or ex-dividend.  If we let T represent the marginal tax rate on dividends, then the marginal tax rate on capital gains is (0.4T).  In order for the net extra return from buying with-dividend (instead of ex-dividend) to be zero:

- Extra investment + After-tax dividend + Reduction in capital gains tax = 0

Therefore, per dollar of dividend:

-0.85 + [(1 - T) ´ (1.00)] + [(0.4T) ´ (0.85)] = 0

T = 0.227 = 22.7%

c.                  We would expect the high-payout stocks to show the largest decline per dollar of dividends paid because these stocks should be held by investors in low, or perhaps even zero, marginal tax brackets.

d.                  Some investors (e.g., pension funds and security dealers) are indifferent between \$1 of dividends and \$1 of capital gains.  These investors should be prepared to buy any amount of stock with-dividend as long as the fall‑off in price is fractionally less than the dividend.  Elton and Gruber’s result suggests that there must be some impediment to such tax arbitrage (e.g., transactions costs or IRS restrictions).  But, in that case, it is difficult to interpret their result as indicative of marginal tax rates.

e.                  Since the passage of the Tax Reform Act, the tax advantage to capital gains has been reduced.  If investors are now indifferent between dividends and capital gains, we would expect that the payment of a \$1 dividend would result in a \$1 decrease in price.

19.       Under the tax system in the United States, the only investors who are indifferent to the dividend payout ratio are those who pay the same tax rate on dividends as on capital gains.  This is true regardless of the corporate tax rate.

Under Australia’s imputation tax system, shareholders pay income tax on dividends received, but they can deduct from their tax bill their share of the corporate tax on pre-tax earnings paid by the company.  The only investors who would be indifferent with regard to the payout ratio are those whose marginal tax rate is 100%, because they do not receive anything after tax, regardless of whether the income is a capital gain or a dividend.  Therefore, all Australian investors prefer dividends because the corporation, in effect, pays part of the personal tax on dividends but pays no part of the personal tax on capital gains.

20.             Even if the middle-of-the-road party is correct about the supply of dividends, we still do not know why investors wanted the dividends they got.  So, it is difficult to be sure about the effect of the tax change.  If there is some non-tax advantage to dividends that offsets the apparent tax disadvantage, then we would expect investors to demand more dividends after the Tax Reform Act.  If the apparent tax disadvantage were irrelevant because there were too many loopholes in the tax system, then the Tax Reform Act would not affect the demand for dividends.  In any case, the middle-of-the-roaders would argue that once companies adjusted the supply of dividends to the new equilibrium, dividend policy would again become irrelevant.

Challenge Questions

1.                  We make use of Lintner’s model, suitably rearranged:

DIVt = Adjustment Rate x Target Ratio x EPS t + (1 – Adjustment Rate) x DIVt  - 1

Thus, if we regress dividends at time t against earnings per share (also at time t) and dividends (at time t – 1), the adjustment rate and the target rate can be found as follows:

Adjustment Rate = 1 – (coefficient of DIVt  - 1)

Target Ratio  = (coefficient of EPS t )/Adjustment Rate

These two regressions were performed using ExcelÒ (forcing the constant to be zero).  The results are:

 Merck International Paper Adjustment Rate 0.043 0.009 Target Ratio 1.574 2.309

For Merck, if EPS in 2001 is \$5, then the predicted dividend in 2001 is:

DIV2001 = (0.043)´(1.574)´(5.00) + (1 - 0.043)´(1.21) = \$1.50

For International Paper, if EPS in 2001 is \$3, then the predicted dividend in 2001 is:

DIV2001 = (0.009)´(2.309)´(3.00) + (1 - 0.009)´(1.00) = \$1

2.                  Reducing the amount of earnings retained each year will, of course, reduce the growth rate of dividends.  Also, the firm will have to issue new shares each year in order to finance company growth.  Under the original dividend policy, we expect next year’s stock price to be: (\$50 ´ 1.08) = \$54.  If N is the number of shares previously outstanding, the value of the company at t = 1 is (54N).

Under the new policy, n new shares will be issued at t = 1 to make up for the reduction in retained earnings resulting from the new policy.  This decrease is:

(\$4 - \$2) = \$2 per original share, or an aggregate reduction of 2N.  If P1 is the price of the common stock at t = 1 under the new policy, then:

2N = nP1

Also, because the total value of the company is unchanged:

54N = (N + n)P1

Solving, we find that P1 = \$52.

If g is the expected growth rate under the new policy and P0 the price at t = 0, we have:

52 = (1 + g)P0

and:

Substituting the second equation above for P0 in the first equation and then solving, we find that g = 4% and P0 = \$50, so that the current stock price is unchanged.

3.                  Generally, a share repurchase is viewed as a signal that::

a.                  Management desires to avoid excess cash, and/or;

b.                  Management desires to Increase the debt:equity ratio, and/or;

c.                  The stock is a good value even at 20% above the current market share price.

Under any or all of these conditions, the share price would likely increase.  Conversely, if the repurchase made the firm substantially more risky, or if managers were having their own shares repurchased, or if the action was interpreted as an inability to find positive NPV projects for the future, then the share price might either remain unchanged or decrease.

4.                  It is true that researchers have been consistent in finding a positive association between price-earnings multiples and payout ratios.  But simple tests like this one do not isolate the effects of dividend policy, so the evidence is not convincing.

Suppose that King Coal Company, which customarily distributes half its earnings, suffers a strike that cuts earnings in half.  The setback is regarded as temporary, however, so management maintains the normal dividend.  The payout ratio for that year turns out to be 100 percent, not 50 percent.

The temporary earnings drop also affects King Coal’s price-earnings ratio.  The stock price may drop because of this year’s disappointing earnings, but it does not drop to one-half its pre-strike value.  Investors recognize the strike as temporary, and the ratio of price to this year’s earnings increases.  Thus, King Coal’s labor troubles create both a high payout ratio and a high price-earnings ratio.  In other words, they create a spurious association between dividend policy and market value.  The same thing happens whenever a firm encounters temporary good fortune, or whenever reported earnings underestimate or overestimate the true long-run earnings on which both dividends and stock prices are based.

A second source of error is omission of other factors affecting both the firm’s dividend policy and its market valuation.  For example, we know that firms seek to maintain stable dividend rates.  Companies whose prospects are uncertain therefore tend to be conservative in their dividend policies.  Investors are also likely to be concerned about such uncertainty, so that the stocks of such companies are likely to sell at low multiples.  Again, the result is an association between the price of the stock and the payout ratio, but it stems from the common association with risk and not from a market preference for dividends.

Another reason that earnings multiples may be different for high-payout and low-payout stocks is that the two groups may have different growth prospects.  Suppose, as has sometimes been suggested, that management is careless in the use of retained earnings but exercises appropriately stringent criteria when spending external funds.  Under such circumstances, investors would be correct to value stocks of high-payout firms more highly.  But the reason would be that the companies have different investment policies.  It would not reflect a preference for high dividends as such, and no company could achieve a lasting improvement in its market value simply by increasing its payout ratio.