DIVIDEND POLICY
At the end of each year, every publicly traded company has to decide whether to return cash to its stockholders and, if yes, how much in the form of dividends. The owner of a private company has to make a similar decision about how much cash he plans to withdraw from the business, and how much toreinvest. This is the dividend decision, and we begin this chapter by providingsome background on th
ree aspects of dividend policy. One is a purely proceduralquestion about how dividends are set and paid out to stockholders. The second isan examination of widely used measures of how much a firm pays in the dividends.The third is an empirical examination of some patterns that firms follow individend policy. Having laid this groundwork, we look at three schools ofthought on dividend policy. The dividend irrelevance school believes thatdividends do not really matter, because they do not affect firm value. Thisargument is based upon two assumptions. The first is that there is no taxdisadvantage to an investor to receiving dividends, and the second is that firmscan raise funds in capital markets for new investments without bearingsignificant issuance costs. The proponents of the second school feel thatdividends are bad for the average stockholder because of the tax disadvantagethey create, which results in lower value. Finally, there are those in a thirdgroup who argue that dividends are clearly good because stockholders (or atleast some of them) like them. Although dividends have traditionally beenconsidered the primary approach for publicly traded firms to return cash orassets to their stockholders, they comprise only one of many ways available tothe firm to accomplish this objective. In particular, firms can return cash tostockholders through equity repurchases, where the cash is used to buy backoutstanding stock in the firm and reduce the number of shares outstanding. Inaddition, firms can return some of their assets to their stockholders in theform of spin offs and split offs. This chapter will focus on dividendsspecifically, but the next chapter will examine the other alternatives availableto firms, and how to choose between dividends and these alternatives. 2 2Background on Dividend Policy In this section, we consider three issues. First,how do firms decide how much to pay in dividends, and how do those dividendsactually get paid to the stockholders? We next consider two widely used measuresof how much a firm pays in dividends, the dividend payout ratio and the dividendyield. We follow up by looking at some empirical evidence on firm behavior insetting and changing dividends. The Dividend Process Firms in the United Statesgenerally pay dividends every quarter, whereas firms in other countriestypically pay dividends on a semi-annual or annual basis. Let us look at thetime line associated with dividend paymen
t and define different types ofdividends. The Dividend Payment Time Line Dividends in publicly traded firms areusually set by the board of directors and paid out to stockholders a few weekslater. There are several key dates between the time the board declares thedividend until the dividend is actually paid.
At the end of each year, every publicly traded company has to decide whether to return cash to its stockholders and, if yes, how much in the form of dividends. The owner of a private company has to make a similar decision about how much cash he plans to withdraw from the business, and how much toreinvest. This is the dividend decision, and we begin this chapter by providingsome background on th
ree aspects of dividend policy. One is a purely proceduralquestion about how dividends are set and paid out to stockholders. The second isan examination of widely used measures of how much a firm pays in the dividends.The third is an empirical examination of some patterns that firms follow individend policy. Having laid this groundwork, we look at three schools ofthought on dividend policy. The dividend irrelevance school believes thatdividends do not really matter, because they do not affect firm value. Thisargument is based upon two assumptions. The first is that there is no taxdisadvantage to an investor to receiving dividends, and the second is that firmscan raise funds in capital markets for new investments without bearingsignificant issuance costs. The proponents of the second school feel thatdividends are bad for the average stockholder because of the tax disadvantagethey create, which results in lower value. Finally, there are those in a thirdgroup who argue that dividends are clearly good because stockholders (or atleast some of them) like them. Although dividends have traditionally beenconsidered the primary approach for publicly traded firms to return cash orassets to their stockholders, they comprise only one of many ways available tothe firm to accomplish this objective. In particular, firms can return cash tostockholders through equity repurchases, where the cash is used to buy backoutstanding stock in the firm and reduce the number of shares outstanding. Inaddition, firms can return some of their assets to their stockholders in theform of spin offs and split offs. This chapter will focus on dividendsspecifically, but the next chapter will examine the other alternatives availableto firms, and how to choose between dividends and these alternatives. 2 2Background on Dividend Policy In this section, we consider three issues. First,how do firms decide how much to pay in dividends, and how do those dividendsactually get paid to the stockholders? We next consider two widely used measuresof how much a firm pays in dividends, the dividend payout ratio and the dividendyield. We follow up by looking at some empirical evidence on firm behavior insetting and changing dividends. The Dividend Process Firms in the United Statesgenerally pay dividends every quarter, whereas firms in other countriestypically pay dividends on a semi-annual or annual basis. Let us look at thetime line associated with dividend paymen
t and define different types ofdividends. The Dividend Payment Time Line Dividends in publicly traded firms areusually set by the board of directors and paid out to stockholders a few weekslater. There are several key dates between the time the board declares thedividend until the dividend is actually paid. • The first date of note is thedividend declaration date, the date on which the board of directors declares thedollar dividend that will be paid for that quarter (or period). This date isimportant because by announcing its intent to increase, decrease, or maintaindividend, the firm conveys information to financial markets. Thus, if the firmchanges its dividends, this is the date on which the market reaction to thechange is most likely to occur.
• The next date of note is the ex-dividend date,at which time investors have to have bought the stock in order to receive thedividend. Since the dividend is not received by investors buying stock after theex-dividend date, the stock price will generally fall on that day to reflectthat loss.
• At the close of the business a few days after the ex-dividend date,the company closes its stock transfer books and makes up a list of theshareholders to date on the holderof- record date. These shareholders willreceive the dividends. There should be generally be no price effect on thisdate.
• The final step involves mailing out the dividend checks on the dividendpayment date. In most cases, the payment date is two to three weeks after theholder-of-record 3 3 date. While stockholders may view this as an important day,there should be no price impact on this day either. Figure 10.1 presents thesekey dates on a time line. Figure 10.1: The Dividend Time Line Board of Directorsannounces quarterly dividend per share Stock has to be bought by this date forinvestor to receive dividends Company closes books and records owners of stockDividend is paid to stockholders Announcement Date Ex-Dividend dayHolder-of-record day Payment day 2 to 3 weeks 2-3 days 2-3 weeks Types ofDividends There are several ways to classify dividends. First, dividends can bepaid in cash or as additional stock. Stock dividends increase the number ofshares outstanding and generally reduce the price per share. Second, thedividend can be a regular dividend, which is paid at regular intervals(quarterly, semi-annually, or annually), or a special dividend, which is paid inaddition to the regular dividend. Most U.S. firms pay regular dividends everyquarter; special dividends are paid at irregular intervals. Finally, firmssometimes pay dividends that are in excess of the retained earnings they show ontheir books. These are called liquidating dividends and are viewed by theInternal Revenue Service as return on capital rather than ordinary income.Consequently, they can have different tax consequences for investors. Measuresof Dividend Policy We generally measure the dividends paid by a firm using oneof two measures. The first is the dividend yield, which relates the dividendpaid to the price of the stock: Dividend Yield = Annual Dividends per share /Price per share The dividend yield is significant because it provides a measureof that component of the total return that comes from dividends, with thebalance coming from price appreciation. Expected Return on Stock = DividendYield + Price Appreciation Some investors also use the dividend yield as ameasure of risk and as an investment screen, i.e., they invest in stocks withhigh dividend yields. Studies indicate that stocks 4 4 with high dividend yieldsearn excess returns, after adjusting for market performance and risk. Figure10.2 tracks dividend yields on the 2700 listed stocks in the United States thatpaid dividends on the major exchanges in January 2004. Note, though, that 4800firms out of the total sample of 7500 firms did not pay dividends. Strictlyspeaking, the median dividend yield for a stock in the United States is zero. aEstimated using Value Line data on companies in January 2004 The median dividendyield among dividend paying stocks is 1.80%, and the average dividend yield of2/12% is low by historical standards, as evidenced by Figure 10.3, which plotsaverage dividend yields by year from 1960 to 2003. Dividend Yield: This is thedollar dividend per share divided by the current price per share. 5 5 aEstimated using S&P 500 data from 1960 to 2003; Source is Bloomberg. The secondwidely used measure of dividend policy is the dividend payout ratio, whichrelates dividends paid to the earnings of the firm. Dividend Payout Ratio =Dividends / Earnings The payout ratio is used in a number of different settings.It is used in valuation as a way of estimating dividends in future periods,since most analysts estimate growth in earnings rather than dividends. Second,the retention ratio –– the proportion of the earnings reinvested in the firm(Retention Ratio = 1 -Dividend Payout Ratio) –– is useful in estimating futuregrowth in earnings; firms with high retention ratios (low payout ratios)generally have higher growth rates in earnings than do firms with lowerretention ratios (higher payout ratios). Third, the dividend payout ratio tendsto follow the life cycle of the firm, starting at zero when the firm is in highgrowth and gradually increasing as the firm matures and its growth prospectsdecrease. Figure 10.4 graphs the dividend payout ratios of U.S. firms that paiddividends in January 2004. Dividend Payout: This is the dividend paid as apercent of the net income of the firm. If the earnings are negative, it is notmeaningful. 6 6 a Estimated using Value Line data on companies in January 2004The payout ratios greater than 100% represent firms that paid out more thantheir earnings as dividends. The median dividend payout ratio in January 2004among dividend paying stocks, was about 30 % while the average payout ratio wasapproximately 35%. 10.1.
☞: Dividends that Exceed Earnings Companies shouldnever pay out more than 100% of their earnings as dividends. a. True b. FalseExplain. divUS.xls: There is a dataset on the web that summarizes dividendyields and payout ratios for U.S. companies from 1960 to the present. 7 7Empirical Evidence on Dividend Policy We observe several interesting patternswhen we look at the dividend policies of firms in the United States in the last50 years. First, dividends tend to lag behind earnings; that is, increases inearnings are followed by increases in dividends, and decreases in earningssometimes by dividend cuts. Second, dividends are “sticky” because firms aretypically reluctant to change dividends; in particular, firms avoid cuttingdividends even when earnings drop. Third, dividends tend to follow a muchsmoother path than do earnings. Finally, there are distinct differences individend policy over the life cycle of a firm, resulting from changes in growthrates, cash flows, and project availability. Dividends Tend To Follow EarningsIt should not come as a surprise that earnings and dividends are positivelycorrelated over time, because dividends are paid out of earnings. Figure 10.5shows the movement in both earnings and dividends between 1960 and 1998. aEstimated from Compustat annual database 8 8 Notice two trends in this graph.First, dividend changes trail earnings changes over time. Second, the dividendseries is much smoother than is the earnings series. In the 1950s, John Lintnerstudied the way firms set dividends and noted three consistent patterns.1 First,firms set target dividend payout ratios, by deciding on the fraction of earningsthey are willing to pay out as dividends in the long term. Second, they changedividends to match long-term and sustainable shifts in earnings, but theyincrease dividends only if they feel they can maintain these higher dividends.Because firms avoid cutting dividends, dividends lag earnings. Finally, managersare much more concerned about changes in dividends than about levels ofdividends. Fama and Babiak identified a lag between earnings and dividends, byregressing changes in dividends against changes in earnings in both current andprior periods2. They confirmed Lintner’s findings that dividend changes tend tofollow earnings changes. 10.2. ☞: Determinants of Dividend Lag Which of thefollowing types of firms is likely to wait least after earnings go up beforeincreasing dividends? a. A cyclical firm, whose earnings have surged because ofan economic boom b. A pharmaceutical firm whose earnings have increased steadilyover the last 5 years, due to a successful new drug. c. A personal computermanufacturer, whose latest laptop’s success has translated into a surge inearnings Explain. Dividends Are Sticky Firms generally do not change theirdollar dividends frequently. This reluctance to change dividends, which resultsin ‘sticky dividends,’ is rooted in several factors. One is 1 Lintner, J.,Distribution of Income of Corporations among Dividends, Retained Earnings andTaxes, American Economic Review, 1956, v46, 97-113. 2 Fama, E. F. and H. Babiak.Dividend Policy: An Empirical Analysis, Journal of the American StatisticalAssociation, 1968, v63(324), 1132-1161. Target Dividend Payout Ratio: This isthe desired proportion of earnings that a firm wants to pay out in dividends. 99 the firm’s concern about its capability to maintain higher dividends in futureperiods. Another is the negative market view of dividend decreases, and theconsequent drop in the stock price. Figure 10.6 provides a summary of thepercentages of all firms that increased, decreased, or left unchanged theirannual dividends per share from 1989 to 1998. a Estimated using Compustat annualdatabase. As you can see, in most years the number of firms that do not changetheir dollar dividends far exceeds the number that do. Among the firms thatchange dividends, a much higher percentage, on average, increase dividends thandecrease them. Dividends Follow a Smoother Path than Earnings As a result of thereluctance of firms to raise dividends until they feel able to maintain them,and to cut dividends unless they absolutely have to, dividends follow a muchsmoother path than earnings. This view that dividends are not as volatile asearnings on a year-to-year basis is supported by a couple of empirical facts.First, the variability in historical dividends is significantly lower than thevariability in historical earnings. Using annual data on aggregate earnings anddividends from 1960 to 2003, for 10 10 instance, the standard deviation ofdividends is 5% while the standard deviation in earnings is about 14%. Second,the standard deviation in earnings yields across companies is significantlyhigher than the standard deviation in dividend yields. In other words, thevariation in earnings yields across firms is much greater than the variation individend yields. A Firm’s Dividend Policy Tends To Follow The Life Cycle Of TheFirm In chapter 7, we introduced the link between a firm’s place in the lifecycle and its financing mix and choices. In particular, we noted five stages inthe growth life cycle – start up, rapid expansion, high growth, mature growthand decline. In this section, we will examine the link between a firm’s place inthe life cycle and its dividend policy. Not surprisingly, firms generally adoptdividend policies that best fit where they are currently in their life cycles.For instance, high-growth firms with great investment opportunities do notusually pay dividends, whereas stable firms with larger cash flows and fewerprojects tend to pay more of their earnings out as dividends. Figure 10.7 looksat the typical path that dividend payout follows over a firm’s life cycle. 11 11Stage 2 Rapid Expansion Stage 1 Start-up Stage 4 Mature Growth Stage 5 DeclineFigure 10.7: Life Cycle Analysis of Dividend Policy Years Years Capacity to paydividends Revenues Earnings None None Increasing High External funding needsHigh, but constrained by infrastructure High, relative to firm value. Moderates,relative to firm value. Low, as projects dry up. Internal financing Low, asprojects dry up. Very low Stage 3 High Growth Negative or low Negative or lowLow, relative to funding needs High, relative to funding needs More than fundingneeds Growth stage $ Revenues/ Earnings This intuitive relationship betweendividend policy and growth is emphasized when we look at the relationshipbetween a firm’s payout ratio and its expected growth rate. For instance, weclassified firms on the New York Stock Exchange in January 2004 into sixclasses, based upon analyst estimates of expected growth rates in earnings pershare for the next 5 years, and estimated the dividend payout ratios anddividend yields for each class; these are reported in Figure 10.8. 12 12 Source:Value Line Database The firms with the highest expected growth rates pay thelowest dividends, both as a percent of earnings (payout ratio) and as a percentof price (dividend yield).3 10.3. ☞: Dividend Policy at Growth Firms Assume thatyou are following a growth firm, whose growth rate has begun easing. Which ofthe following would you most likely observe in terms of dividend policy at thefirm? a. An immediate increase of dividends to reflect the lower reinvestmentneeds. b. No change in dividend policy, and an increase in the cash balance. c.No change in dividend policy, and an increase in acquisitions of other firmsExplain. 3 These are growth rates projected by Value Line for firms in April1999. 13 13 Differences in Dividend Policy across Countries Figures 10.5 to 10.8showed several trends and patterns in dividend policies at U.S. companies.4 Theyshare some common features with firms in other countries, and there are somedifferences. As in the United States, dividends in other countries are stickyand follow earnings. However, there are differences in the magnitude of dividendpayout ratios across countries. Figure 10.9 shows the proportion of earningspaid out in dividends in the G-7 countries in 1982-84 and again in 1989-91.aSource: Rajan and Zingales These differences can be attributed to: 1.Differences in Stage of Growth: Just as higher growth companies tend to pay outless of their earnings in dividends (see Figure 10.8), countries with highergrowth pay out less in dividends. For instance, Japan had much higher expectedgrowth in 1982-84 than the other G-7 countries and paid out a much smallerpercentage of its earnings as dividends. 2. Differences in Tax Treatment: Unlikethe United States, where dividends are double taxed, some countries provide atleast partial protection against the double taxation of 4 Rajan, R. and L.Zingales, What do we know about capital structure? Some Evidence fromInternational Data, Journal of Finance, 1995, v50, 1421-1460. 14 14 dividends.For instance, Germany taxes corporate retained earnings at a higher rate thancorporate dividends. 3. Differences in Corporate Control: When there is aseparation between ownership and management, as there is in many large publiclytraded firms, and where stockholders have little control over managers, thedividends paid by firms will be lower. Managers, left to their own devices, havea much greater incentive to accumulate cash than do stockholders. Notsurprisingly, the dividend payout ratios of companies in emerging markets aremuch lower than the dividend payout ratios in the G-7 countries. The highergrowth and relative power of incumbent management in these countries contributeto keeping these payout ratios low. 10.4.
☞: Dividend policies and Stock BuybackRestrictions Some countries do not allow firms to buy back stock from theirstockholders. Which of the following would you expect of dividend policies inthese countries (relative to countries that don’t restrict stock buybacks)? a.Higher portion of earnings will be paid out in dividends; More volatiledividends; b. Lower portion of earnings will be paid out in dividends; Morevolatile dividends c. Higher portion of earnings will be paid out in dividends;Less volatile dividends; d. Lower portion of earnings will be paid out individends; Less volatile dividends; Explain Illustration 10.1: Dividends,Dividend Yields and Payout Ratios In the illustration that follows, we willexamine the dollar dividends paid at Disney, Aracruz and Deutsche Bank between2001 and 2003. Each year, we will also compute the dividend yield and dividendpayout ratio for each firm. Deutsche Bank Disney Aracruz 2001 2002 2003 20012002 2003 2001 2002 2003 DPS € 1.30 € 1.30 € 1.30 $0.21 $0.21 $0.21 R$ 0.14 R$0.18 R$ 0.32 EPS € 2.44 € 0.64 € 0.27 $0.11 $0.60 $0.65 R$ 0.20 R$ 0.01 R$ 0.85Stock Price € 79.40 € 43.90 € 65.70 $20.72 $16.31 $23.33 R$ 3.91 R$ 6.76 R$10.60 Dividend Yield 1.64% 2.96% 1.98% 1.01% 1.29% 0.90% 3.53% 2.69% 3.00%Dividend Payout 53.28% 203.13% 481.48% 190.91% 35.00% 32.31% 69.01% 1818.44%37.41% 15 15 Of the three companies, Aracruz had the highest dividend yieldacross the three years. Disney and Deutsche paid the same dividends per shareeach year, but the volatility in their stock prices and earnings made the payoutratios and dividend yields volatile. In fact, Deutsche maintained its dividendsat 1.30 Euros per share in the face of declining earnings per share in 2002 and2003, a testimonial to the stickiness of dividends. As noted earlier in thebook, Aracruz, like most Brazilian companies, maintains two classes of shares –voting share (called common and held by insiders) and non-voting shares (calledpreferred shares and held by outside investors). The dividend policies aredifferent for the two classes, with preferred shares getting higher dividends.In fact, the failure to pay a mandated dividend to preferred stockholders(usually set at a payout ratio of 35%) can result in preferred stockholdersgetting some voting control of the firm. Effectively, this puts a floor on thedividend payout ratio. When Are Dividends Irrelevant? There is a school ofthought that argues that what a firm pays in dividends is irrelevant and thatstockholders are indifferent about receiving dividends. Like the capitalstructure irrelevance proposition, the dividend irrelevance argument has itsroots in a paper crafted by Miller and Modigliani.5 The Underlying AssumptionsThe underlying intuition for the dividend irrelevance proposition is simple.Firms that pay more dividends offer less price appreciation but must provide thesame total return to stockholders, given their risk characteristics and the cashflows from their investment decisions. Thus, there are no taxes, or if dividendsand capital gains are taxed at the same rate, investors should be indifferent toreceiving their returns in dividends or price appreciation. For this argument towork, in addition to assuming that there is no tax advantage or disadvantageassociated with dividends, we also have to assume the following: 5 Miller, M.and F. Modigliani, 1961, Dividend Policy, Growth and the Valuation of Shares,Journal of Business, 411-433. 16 16
• There are no transactions costs associatedwith converting price appreciation into cash, by selling stock. If this were nottrue, investors who need cash urgently might prefer to receive dividends.
•Firms that pay too much in dividends can issue stock, again with no flotation ortransactions costs, to take on good projects. There is also an implicitassumption that this stock is fairly priced.
• The investment decisions of thefirm are unaffected by its dividend decisions, and the firm’s operating cashflows are the same no matter which dividend policy is adopted.
• Managers offirms that pay too little in dividends do not waste the cash pursuing their owninterests (i.e., managers with large free cash flows do not use them to take onbad projects). Under these assumptions, neither the firms paying the dividendsnor the stockholders receiving them will be adversely affected by firms payingeither too little or too much in dividends. 10.5
. ☞: Dividend Irrelevance Basedupon the Miller Modigliani assumptions, dividends are least likely to affectvalue for the following types of firms a. Small companies with substantialinvestment needs. b. Large companies with significant insider holdings. c. Largecompanies with significant holdings by pension funds (which are tax exempt) andminimal investment needs. Explain. A Proof of Dividend Irrelevance To provide aformal proof of irrelevance, assume that LongLast Corporation, an unlevered firmmanufacturing furniture, has operating income after taxes of $ 100 million,growing at 5% a year, and that its cost of capital is 10%. Further, assume thatthis firm has reinvestment needs of $ 50 million, also growing at 5% a year, andthat there are 105 million shares outstanding. Finally, assume that this firmpays out residual cash flows as dividends each year. The value of LongLastCorporation can be estimated as follows: 17 17 Free Cash Flow to the Firm = EBIT(1- tax rate) – Reinvestment needs = $ 100 million - $ 50 million = $ 50 millionValue of the Firm = Free Cash Flow to Firm (1+g) / (WACC - g) = $ 50 (1.05) /(.10 - .05) = $ 1050 million Price per share = $ 1050 million / 105 million = $10.00 Based upon its cash flows, this firm could pay out $ 50 million individends. Dividend per share = $ 50 million/105 million = $ 0.476 Total Valueper Share = $ 10.00 + $ 0.48 = $10.476 The total value per share measures whatstockholders gets in price and dividends from their stock holdings. Scenario 1:LongLast doubles dividends To examine how the dividend policy affects firmvalue, assume that LongLast Corporation is told by an investment banker that itsstockholders would gain if the firm paid out $ 100 million in dividends, insteadof $ 50 million. It now has to raise $ 50 million in new financing to cover itsreinvestment needs. Assume that LongLast Corporation can issue new stock with noissuance cost to raise these funds. If it does so, the firm value will remainunchanged, since the value is determined not by the dividend paid but by thecash flows generated on the projects. Since the growth rate and the cost ofcapital are unaffected, we get: Value of the Firm = $ 50 (1.05) / (.10 - .05) =$ 1050 million The existing stockholders will receive a much larger dividend pershare, since dividends have been doubled: Dividends per share = $ 100million/105 million shares = $ 0.953 In order to estimate the price per share atwhich the new stock will be issued, note that after the new stock issue of $ 50million, the old stockholders in the firm will own only $1000 million of thetotal firm value of $ 1050 million. Value of the Firm for existing stockholdersafter dividend payment = $ 1000 million Price per share = $ 1000 million / 105million = $ 9.523 The price per share is now lower than it was before thedividend increase, but it is exactly offset by the increase in dividends. Valueaccruing to stockholder = $ 9.523 + $ 0.953 = $ 10.476 18 18 Thus, if theoperating cash flows are unaffected by dividend policy, we can show that thefirm value will be unaffected by dividend policy and that the averagestockholder will be indifferent to dividend policy, since he or she receives thesame total value (price + dividends) under any dividend payment. Scenario 2:LongLast stops paying dividends To consider an alternate scenario, assume thatLongLast Corporation pays out no dividends and retains the residual $50 millionas a cash balance. The value of the firm to existing stockholders can then becomputed as follows: Value of Firm = Present Value of After-tax Operating CF +Cash Balance = $ 50 (1.05) / (.10 - .05) + $ 50 million = $1100 million Valueper share = $ 1100 million / 105 million shares = $10.48 Note that the totalvalue per share remains at % 10.48. In fact, as shown in Table 10.1, the valueper share remains $10.48, no matter how much the firm pays in dividends. Table10.1: Value Per Share to Existing Stockholders from Different Dividend PoliciesValue of Firm Dividends Value to Existing Price Dividends Total Value (OperatingCF) Stockholders per share per share per share $1,050 $ - $1,100 $ 10.48 $ - $10.48 $1,050 $ 10.00 $1,090 $ 10.38 $ 0.10 $ 10.48 $1,050 $ 20.00 $1,080 $ 10.29$ 0.19 $ 10.48 $1,050 $ 30.00 $1,070 $ 10.19 $ 0.29 $ 10.48 $1,050 $ 40.00$1,060 $ 10.10 $ 0.38 $ 10.48 $1,050 $ 50.00 $1,050 $ 10.00 $ 0.48 $ 10.48$1,050 $ 60.00 $1,040 $ 9.90 $ 0.57 $ 10.48 $1,050 $ 70.00 $1,030 $ 9.81 $ 0.67$ 10.48 $1,050 $ 80.00 $1,020 $ 9.71 $ 0.76 $ 10.48 $1,050 $ 90.00 $1,010 $ 9.62$ 0.86 $ 10.48 $1,050 $ 100.00 $1,000 $ 9.52 $ 0.95 $ 10.48 When LongLastCorporation pays less than $ 50 million in dividends, the cash accrues in thefirm and adds to its value. The increase in the stock price again is offset bythe loss of 19 19 cash flows from dividends. When it pays out more, the pricedecreases but is exactly offset by the increase in dividends per share. Note,though, that the value per share remains unchanged because we assume that thereare no tax differences to investors between dividends and capital gains, thatfirms can raise new capital with no issuance costs, and that firms do not changetheir investment policy. These assumptions eliminate the costs associated withpaying either more in dividends or less. Implications of Dividend Irrelevance Ifdividends are, in fact, irrelevant, firms are spending a great deal of timepondering an issue about which their stockholders are indifferent. A number ofstrong implications emerge from this proposition. Among them, the value ofequity in a firm should not change as its dividend policy changes. This does notimply that the price per share will be unaffected, however, since largerdividends should result in lower stock prices and more shares outstanding. Inaddition, in the long term, there should be no correlation between dividendpolicy and stock returns. Later in this chapter, we will examine some studiesthat have attempted to examine whether dividend policy is in fact irrelevant inpractice. The assumptions needed to arrive at the dividend irrelevanceproposition may seem so onerous that many reject it without testing it. Thatwould be a mistake, however, because the argument does contain a valuablemessage: Namely, a firm that has invested in bad projects cannot hope toresurrect its image with stockholders by offering them higher dividends. Infact, the correlation between dividend policy and total stock returns is weak,as we will see later in this chapter. The “Dividends Are Bad” School In theUnited States, dividends have historically been taxed at much higher rates thancapital gains. Based upon this tax disadvantage, the second school of thought ondividends argued that dividend payments reduce the returns to stockholders afterpersonal taxes. Stockholders, they posited, would respond by reducing the stockprices of the firms making these payments, relative to firms that do not paydividends. Consequently, firms would be better off either retaining the moneythey would have paid out as dividends or 20 20 repurchasing stock. In 2003, thebasis for this argument was largely eliminated when the tax rate on dividendswas reduced to match the tax rate on capital gains. In this section, we willconsider both the history of tax-disadvantaged dividends and the potentialeffects of the tax law changes.6 The History of Dividend Taxation The taxtreatment of dividends varies widely depending upon who receives the dividend.Individual investors are taxed at ordinary tax rates, corporations are shelteredfrom paying taxes on at least a portion of the dividends they receive andpension funds are not taxed at all. Individuals Since the inception of incometaxes in the early part of the twentieth century in the United States, dividendsreceived on investments have been treated as ordinary income, when received byindividuals, and taxed at ordinary tax rates. In contrast, the priceappreciation on an investment has been treated as capital gains and taxed at adifferent and much lower rate. Figure 10.10 graphs the highest marginal tax rateon dividends in the United States and the highest marginal capital gains taxrate since 1954 (when capital gains taxes were introduced). 6 Adding to theuncertainty is the fact that the tax changes of 2003 are not permanent and aredesigned to sunset (disappear) in 2010. It is unclear whether the taxdisadvantages of dividends have disappeared for the long term or only until2010. 21 21 Barring a brief period after the 1986 tax reform act, when dividendsand capital gains were both taxed at 28%, the capital gains tax rate has beensignificantly lower than the ordinary tax rate in the United States. In 2003,the tax rate on dividends was dropped to 15% to match the tax rate on capitalgains, thus nullifying the tax disadvantage of dividends. There are two pointsworth making about this chart. The first is that these are the highest marginaltax rates and that most individuals are taxed at lower rates. In fact, someolder and poorer investors may pay no taxes on income, if their income fallsbelow the threshold for taxes. The second and related issue is that the capitalgains taxes can be higher for some of these individuals than the ordinary taxrate they pay on dividends. Overall, though, wealthier individuals have moreinvested in stocks than poorer individuals, and it seems fair to conclude thatindividuals have collectively paid significant taxes on the income that theyhave received in dividends over the last few decades. 22 22 InstitutionalInvestors About two-thirds of all traded equities are held by institutionalinvestors rather than individuals. These institutions include mutual funds,pension funds and corporations and dividends get taxed differently in the handsof each. • Pension funds are tax-exempt. They are allowed to accumulate bothdividends and capital gains without having to pay taxes. There are two reasonsfor this tax treatment. One is to encourage individuals to save for theirretirement and to reward savings (as opposed to consumption). The other reasonfor this is that individuals will be taxed on the income they receive from theirpension plans and that taxing pension plans would in effect tax the same incometwice. • Mutual funds are not directly taxed, but investors in mutual funds aretaxed for their share of the dividends and capital gains generated by the funds.If high tax rate individuals invest in a mutual fund that invests in stocks thatpay high dividends, these high dividends will be allocated to the individualsbased on their holdings and taxed at their individual tax rates. • Corporationsare given special protection from taxation on dividends they receive on theirholdings in other companies, with 70% of the dividends exempt from taxes7. Inother words, a corporation with a 40% tax rate that receives $ 100 million individends will pay only $12 million in taxes. Here again, the reasoning is thatdividends paid by these corporations to their stockholders will ultimately betaxed. Tax Treatment of Dividends in other markets Many countries have plans inplace to protect investors from the double taxation of divided. There are twoways in which they can do this. One is to allow corporations to claim a full orpartial tax deduction for dividends paid. The other is to give partial or fulltax relief to individuals who receive dividends. 7 The exemption increases asthe proportion of the stock held increases. Thus, a corporation that owns 10% ofanother company’s stock has 70% of dividends exempted. This rises to 80% if thecompany owns between 20 and 80% of the stock and to 100% if the company holdsmore than 80% of the outstanding stock. 23 23 Corporate Tax Relief In somecountries, corporations are allowed to claim a partial or full deduction fordividends paid. This brings their treatment into parity with the treatment ofthe interest paid on debt, which is entitled to a full deduction in mostcountries. Among the OECD countries, the Czech Republic and Iceland offerpartial deductions for dividend payments made by companies but no country allowsa full deduction. In a variation, Germany, until recently, applied a higher taxrate to income that was retained by firms than to income that was paid out individends. In effect, this gives a partial tax deduction to dividends. Why don’tmore countries offer tax relief to corporations? There may be two factors. Oneis the presence of foreign investors in the stock who now also share in the taxwindfall. The other is that investors in the stock may be tax exempt or pay notaxes, which effectively reduces the overall taxes paid on dividends to thetreasury to zero. Individual Tax Relief There are far more countries that offertax relief to individuals than to corporations. This tax relief can take severalforms: • Tax Credit for taxes paid by corporation: Individuals can be allowed toclaim the taxes paid by the corporation as a tax credit when computing their owntaxes. In the example earlier in the paper, where a company paid 30% of itsincome of $ 100 million as taxes and then paid its entire income as dividends toindividuals with 40% tax rates the individuals would be allowed to claim a taxcredit of $ 30 million against the taxes owed, thus reducing taxes paid to $ 10million. In effect, this will mean that only individuals with marginal tax ratesthat exceed the corporate tax rate will be taxed on dividends. Australia,Finland, Mexico, Australia and New Zealand allow individuals to get a fullcredit for corporate taxes paid. Canada, France, the U.K and Turkey allow forpartial tax credits. • Lower Tax Rate on dividends: Dividends get taxed at alower rate than other income to reflect the fact that it is paid out ofafter-tax income. In some countries, the tax rate on dividends is set equal tothe capital gains tax rate. Korea, for instance, has a flat tax rate of 16.5%for dividend income. 24 24 In summary, it is far more common for countries toprovide tax relief to investors than to corporations. Part of the reason forthis is political. By focusing on individuals, you can direct the tax reliefonly towards domestic investors and only to those investors who pay taxes in thefirst place. Timing of Tax Payments When the 1986 tax law was signed into law,equalizing tax rates on ordinary income and capital gains, some believed thatall the tax disadvantages of dividends had disappeared. Others noted that, evenwith the same tax rates, dividends carried a tax disadvantage because theinvestor had no choice as to when to report the dividend as income; taxes weredue when the firm paid out the dividends. In contrast, investors retaineddiscretionary power over when to recognize and pay taxes on capital gains, sincesuch taxes were not due until the stock was sold. This timing option allowed theinvestor to reduce the tax liability in one of two ways. First, by takingcapital gains in periods of low income or capital losses to offset against thegain, the investor could now reduce the taxes paid. Second, deferring a stocksale until an investor’s death could result in tax savings. Since the tax rateson capital gains have decreased relative to the tax rates on dividends since,this timing option should make capital gains an even more attractive option now.Assessing Investor tax preferences for dividends As you can see from thediscussion above, the tax rate on dividends can vary widely for differentinvestors – individual, pension fund, mutual fund or corporation – receiving thedividends and even for the same investor on different investments. It isdifficult therefore to look at a company’s investor base and determine theirpreferences for dividends and capital gains. A simple way to measure the taxdisadvantage associated with dividends is to measure the price change on theex-dividend date and compare it to the actual dividend paid. The stock price onthe ex-dividend day should drop to reflect the loss in dividends to those buyingthe stock after that day. It is not clear, however, whether the price drop willbe equal to the dividends if dividends and capital gains are taxed at differentrates. 25 25 To see the relationship between the price drop and the tax rates ofthe marginal investor, assume that investors in a firm acquired stock at somepoint in time at a price P, and that they are approaching an ex-dividend day, inwhich the dividend is known to be D. Assume that each investor in this firm caneither sell the stock before the ex-dividend day at a price PB or wait and sellit after the stock goes ex-dividend at a price PA. Finally, assume that the taxrate on dividends is to and that the tax rate on capital gains is tcg. The cashflows the investor will receive from selling before the stock goes ex-dividendis – CFB = PB - (PB - P) tcg In this case, by selling before the ex-dividendday, the investor receives no dividend. If the sale occurs after the ex-dividendday, the cash flow is – CFA = PA - (PA - P) tcg + D (1-to) If the cash flow fromselling before the ex-dividend day were greater than the cash flow from sellingafter, the investors would all sell before, resulting in a drop in the stockprice. Similarly, if the cash flows from selling after the ex-dividend day weregreater than the cash flows from selling before, every one would sell after,resulting in a price drop after the ex-dividend day. To prevent either scenario,the marginal investors in the stock have to be indifferent between sellingbefore and after the ex-dividend day. This will occur only if the cash flowsfrom selling before are equal to the cash flows from selling after: PB - (PB -P) tcg = PA - (PA - P) tcg + D (1-to) This can be simplified to yield thefollowing ex-dividend day equality: PB ! PA D = (1- t o) (1 ! tcg) Thus, anecessary condition for the marginal investor to be indifferent between sellingbefore and after the ex-dividend day is that the price drop on the ex-dividendday must reflect the investor’s tax differential between dividends and capitalgains. Turning this equation around, we would argue that by observing a firm’sstock price behavior on the ex-dividend day and relating it to the dividendspaid by the firm, we can, in the long term, form some conclusions about the taxdisadvantage the firm’s stockholders attach to dividends. In particular: If TaxTreatment of Dividends and Capital Gains 26 26 PB - PA = D Marginal investor isindifferent between dividends and capital gains PB - PA <> D Marginal investoris taxed more heavily on capital gains While there are obvious measurementproblems associated with this measure, it does provide some interesting insightinto how investors view dividends. The first study of ex-dividend day pricebehavior was completed by Elton and Gruber in 1970.8 They examined the behaviorof stock prices on ex-dividend days for stocks listed on the NYSE between 1966and 1969. Based upon their finding that the price drop was only 78% of thedividends paid, Elton and Gruber concluded that dividends are taxed more heavilythan capital gains. They also estimated the price change as a proportion of thedividend paid for firms in different dividend yield classes and reported thatprice drop is larger, relative to the dividend paid, for firms in the highestdividend yield classes than for firms in lower dividend yield classes. Thisdifference is price drops, they argued, reflected the fact that investors inthese firms are in lower tax brackets. Their conclusions were challenged,however, by some who argued, justifiably, that the investors trading on thestock on ex-dividend days are not the normal investors in the firm; rather, theyare short-term, tax-exempt investors interested in capturing the differencebetween dividends and the price drops. Implications There can be no argumentthat dividends have historically been treated less favorably than capital gainsby the tax authorities. In the United States, the double taxation of dividends,at least at the level of individual investors, should have created a strongdisincentive to pay or to increase dividends. Other implications of the taxdisadvantage argument include the following: • Firms with an investor basecomposed primarily of individuals typically should have paid lower dividendsthan do firms with investor bases predominantly made up of tax-exemptinstitutions. 8 Elton, E.J. and M.J. Gruber, 1970, Marginal Stockholder Ratesand the Clientele Effect, Review of Economics and Statistics, v52, 68-74. 27 27• The higher the income level (and hence the tax rates) of the investors holdingstock in a firm, the lower the dividend paid out by the firm. • As the taxdisadvantage associated with dividends increased, the aggregate amount paid individends should have decreased. Conversely, if the tax disadvantage associatedwith dividends decreased, the aggregate amount paid in dividends should haveincreased. The tax law changes of 2003 have clearly changed the terms of thisdebate. By reducing the tax rate on dividends, they have clearly made dividendsmore attractive at least to individual investors than they were prior to thechange. We would expect companies to pay more dividends in response. While it isstill early to test out this hypothesis, there is some evidence that companiesare changing dividend policy in response to the tax law change. Technologycompanies like Microsoft that have never paid dividends before have initiateddividends. In figure 10.11, we look at the percent of S&P 500 companies that paydividends by year and the dividends paid as a percent of the marketcapitalization of these companies from 1960 to 2003. 28 28 There was an up tickin both the number of companies paying dividends in 2003 and the dividends paid,reversing a long decline in both statistics. It will be interesting to seewhether this continues into the future. In Practice: Dividend Policy for thenext decade As firms shift towards higher dividends, they may be put at riskbecause of volatile earnings. There are two ways in which they can alleviate theproblem. • One is to shift to a policy of residual dividends, where dividendspaid are a function of the earnings in the year rather than a function ofdividends last year. Note that the sticky dividend phenomenon in the US, wherecompanies are reluctant to change their dollar dividends, is not a universalone. In countries like Brazil, companies target dividend payout ratios ratherthan dollar dividends and there is no reason why US companies cannot adopt asimilar practice. A firm that targets a constant dividend payout ratio will paymore dividends when its earnings are high and less when its earnings are low,and the signaling effect of lower dividends will be mitigated if the payoutpolicy is clearly stated up front. • The other is to adopt a policy of regulardividends that will be based upon sustainable and predictable earnings and tosupplement these with special dividends when earnings are high. In this form,the special dividends will take the place of stock buybacks. In summary, you canexpect both more dividends from companies and more creative dividend policies,if the dividend tax law stands. British Petroleum provided a preview ofinnovations to come by announcing that they would supplement their regulardividends with any extra cashflows generated if the oil price stayed above $ 30a barrel, thus creating dividends that are tied more closely to their cashflows.10.6. ☞: Corporate Tax Status and Dividend Policy Corporations are exempt frompaying taxes on 70% of the dividends they receive from their stockholdings inother companies, whereas they face a capital gains tax rate of 20%. If all thestock in your company is held by other companies, and the ordinary tax rate forcompanies is 36%, a. dividends have a tax advantage relative to capital gains 2929 b. capital gains have a tax advantage relative to dividends c. dividends andcapital gains are taxed at the same rate Explain. The “Dividends Are Good”School Notwithstanding the tax disadvantages, firms continue to pay dividendsand they typically view such payments positively. A third school of thought thatargues dividends are good and can increase firm value. Some of the argumentsused by this school are questionable, but some have a reasonable basis in fact.We consider both in this section. Some Reasons for Paying Dividends that do notmeasure up Some firms pay and increase dividends for the wrong reasons. We willconsider two of those reasons in this section. The Bird-in-the-Hand Fallacy Onereason given for the view that investors prefer dividends to capital gains isthat dividends are certain, whereas capital gains are uncertain. Proponents ofthis view of dividend policy feel that risk averse investors will thereforeprefer the former. This argument is flawed. The simplest counter-response is topoint out that the choice is not between certain dividends today and uncertaincapital gains at some unspecified point in the future, but between dividendstoday and an almost equivalent amount in price appreciation today. Thiscomparison follows from our earlier discussion, where we noted that the stockprice dropped by slightly less than the dividend on the ex-dividend day. Bypaying the dividend, the firm causes its stock price to drop today. Anotherresponse to this argument is that a firm’s value is determined by the cash flowsfrom its projects. If a firm increases its dividends but its investment policyremains unchanged, it will have to replace the dividends with new stock issues.The investor who receives the higher dividend will therefore find himself orherself losing, in present value terms, an equivalent amount in priceappreciation. 30 30 Temporary Excess Cash In some cases, firms are tempted topay or initiate dividends in years in which their operations generate excesscash. Although it is perfectly legitimate to return excess cash to stockholders,firms should also consider their own long-term investment needs. If the excesscash is a temporary phenomenon, resulting from having an unusually good year ora non-recurring action (such as the sale of an asset), and the firm expects cashshortfalls in future years, it may be better off retaining the cash to coversome or all these shortfalls. Another option is to pay the excess cash as adividend in the current year and issue new stock when the cash shortfall occurs.This is not very practical because the substantial expense associated with newsecurity issues makes this a costly strategy in the long term. Figure 10.12summarizes the cost of issuing bonds and common stock, by size of issue in theUnited States.9 Source: Ibbotson, Sindelar and Ritter (1997) 9 Ibbotson, R. G.,J. L. Sindelar and J. R. Ritter. 1988, Initial Public Offerings, Journal ofApplied Corporate Finance, v1(2), 37-45. 31 31 Since issuance costs increase asthe size of the issue decreases and for common stock issues, small firms shouldbe especially cautious about paying out temporary excess cash as dividends. Thissaid, it is important to note that some companies do pay dividends and issuestock during the course of the same period, mostly out of a desire to maintaintheir dividends. Figure 10.13 reports new stock issues by firms as a percentageof firm value, classified by their dividend yields, in 1998. Figure 21.11:Equity Issues by Dividend Class, United States - 1998 0.0% 1.0% 2.0% 3.0% 4.0%5.0% 6.0% 7.0% No dividends <> 4.5% Dividend Yield Class Equity Issues as % ofMarket Value 0.00% 10.00% 20.00% 30.00% 40.00% 50.00% 60.00% Percent of Firmsmaking Equity Issues New Issue Yld New Issues Source: Compustat database, 1998While it is not surprising that stocks that pay no dividends are most likely toissue stock, it is surprising that firms in the highest dividend yield classalso issue significant proportions of new stock (approximately half of all thefirms in this class also make new stock issues). This suggests that many ofthese firms are paying dividends on the one hand and issuing stock on the other,creating significant issuance costs for their stockholders in the process. SomeGood Reasons for Paying Dividends While the tax disadvantages of dividends areclear, especially for individual investors, there are some good reasons whyfirms that are paying dividends should not 32 32 suspend them. First, there areinvestors who like to receive dividends, either because they pay no or very lowtaxes, or because they need the regular cash flows. Firms that have paiddividends over long periods are likely to have accumulated investors with thesecharacteristics, and cutting or eliminating dividends would not be viewedfavorably by this group. Second, changes in dividends allow firms to signal tofinancial markets how confident they feel about future cash flows. Firms thatare more confident about their future are therefore more likely to raisedividends; stock prices often increase in response. Cutting dividends is viewedby markets as a negative signal about future cashflows, and stock prices oftendecline in response. Third, firms can use dividends as a tool for altering theirfinancing mix and moving closer to an optimal debt ratio. Finally, thecommitment to pay dividends can help reduce the conflicts between stockholdersand managers, by reducing the cash flows available to managers. Some investorslike dividends Many in the “dividends are bad” school of thought argue thatrational investors should reject dividends due to their tax disadvantage.Whatever you might think of the merits of that argument, some investors have astrong preference for dividends and view large dividends positively. The moststriking empirical evidence for this comes from studies of companies that havetwo classes of shares: one that pays cash dividends, and another that pays anequivalent amount of stock dividends; thus, investors are given a choice betweendividends and capital gains. John Long studied the price differential on Class Aand B shares traded on Citizens Utility.10 Class B shares paid a cash dividend,while Class A shares paid an equivalent stock dividend. Moreover, Class A sharescould be converted at little or no cost to Class A shares at the option of itsstockholders. Thus, an investor could choose to buy Class B shares to get cashdividends, or Class A shares to get an equivalent capital gain. During theperiod of this study, the tax advantage was clearly on the side of capitalgains; thus, we would expect to find Class B shares selling at a discount onClass A 10 Long, John B., Jr., 1978, The Market Valuation Of Cash Dividends: ACase To Consider, Journal of Financial Economics, 1978, v6(2/3), 235-264. 33 33shares. The study found, surprisingly, that the Class B shares sold at a premiumover Class A shares. Figure 10.14 reports the price differential between the twoshare classes over the period of the analysis. Figure 10.14: Price Differentialon Citizen’s Utility Stock Source: Long (1978) While it may be tempting toattribute this phenomenon to the irrational behavior of investors, such is notthe case. Not all investors like dividends –– many feel its tax burden–– butthere are also many who view dividends positively. These investors may not bepaying much in taxes and consequently do not care about the tax disadvantageassociated with dividends. Or they might need and value the cash flow generatedby the dividend payment. Why, you might ask, do they not sell stock to raise thecash flow they need? The transactions costs and the difficulty of breaking upsmall holdings11 and selling unit shares may make selling small amounts of stockinfeasible. 11 Consider a stockholder who owns 100 shares trading at $ 20 pershare, on which she receives a dividend of $0.50 per share. If the firm did notpay a dividend, the stockholder would have to sell 2.5 shares of stock to raisethe $ 5 that would have come from the dividend. 34 34 Bailey extended Long’sstudy to examine Canadian utility companies, which also offered dividend andcapital-gains shares, and had similar findings.12 Table 10.2 summarizes theprice premium at which the dividend shares sold. Table 10.2: Price Differentialbetween Cash and Stock Dividend Shares Company Premium on Cash Dividend Sharesover Stock Dividend Shares Consolidated Bathurst 19.30% Donfasco 13.30% DomePetroleum 0.30% Imperial Oil 12.10% Newfoundland Light & Power 1.80% RoyalTrustco 17.30% Stelco 2.70% TransAlta 1.10% Average 7.54% Source: Bailey (1988)Note, once again, that on average the cash dividend shares sell at a premium of7.5% over the stock dividend shares. We caution that while these findings do notindicate that all stockholders like dividends, they do indicate that thestockholders in these specific companies liked cash dividends so much that theywere willing to overlook the tax disadvantage and pay a premium for shares thatoffered them. The Clientele Effect Stockholders examined in the studiesdescribed above clearly like cash dividends. At the other extreme are companiesthat pay no dividends, such as Microsoft, and whose stockholders seem perfectlycontent with that policy. Given the vast diversity of stockholders, it is notsurprising that, over time, stockholders tend to invest in firms whose dividendpolicies match their preferences. Stockholders in high tax brackets who do notneed the cash flow from dividend payments tend to invest in companies that paylow or no dividends. By contrast, stockholders in low tax brackets who need thecash from dividend payments, and tax-exempt institutions that need current cashflows, will usually invest in companies with high dividends. This clustering ofstockholders in 12 Bailey, W., 1988, Canada's Dual Class Shares: FurtherEvidence On The Market Value Of Cash Dividends, Journal of Finance, 1988,v43(5), 1143-1160 35 35 companies with dividend policies that match theirpreferences is called the clientele effect. The existence of a clientele effectis supported by empirical evidence. One study looked at the portfolios of 914investors to see whether their portfolios were affected by their tax brackets.The study found that older and poorer investors were more likely to holdhigh-dividend-paying stocks than were younger and wealthier investors. Inanother study, dividend yields were regressed against the characteristics of theinvestor base of a company (including age, income and differential tax rates).13Dividend Yieldt = a + b βt + c Aget + d Incomet + e Differential Tax Ratet + εtVariable Coefficient Implies Constant 4.22% Beta Coefficient -2.145 Higher betastocks pay lower dividends. Age/100 3.131 Firms with older investors pay higherdividends. Income/1000 -3.726 Firms with wealthier investors pay lowerdividends. Differential Tax Rate -2.849 If ordinary income is taxed at a higherrate than capital gains, the firm pays less dividends. Source: Pettit (1977) Notsurprisingly, this study found that safer companies, with older and poorerinvestors, tended to pay more in dividends than companies with wealthier andyounger investors. Overall, dividend yields decreased as the tax disadvantage ofdividends increased. 10.7. ☞: Dividend Clientele and Tax Exempt InvestorsPension funds are exempt from paying taxes on either ordinary income or capitalgains, and also have substantial ongoing cash flow needs. What types of stockswould you expect these funds to buy? a. Stocks that pay high dividends b. Stocksthat pay no or low dividends Explain. 13 Pettit, R. R., 1977, Taxes,Transactions Costs and the Clientele Effect of Dividends, Journal of FinancialEconomics, v5, 419-436. 36 36 Consequences of the Clientele Effect The existenceof a clientele effect has some important implications. First, it suggests thatfirms get the investors they deserve, since the dividend policy of a firmattracts investors who like it. Second, it means that firms will have adifficult time changing an established dividend policy, even if it makescomplete sense to do so. For instance, U.S. telephone companies havetraditionally paid high dividends and acquired an investor base that liked thesedividends. In the 1990s, many of these firms entered new businesses(entertainment, multi-media etc.), with much larger reinvestment needs and lessstable cash flows. While the need to cut dividends in the face of the changingbusiness mix might seem obvious, it was nevertheless a hard sell tostockholders, who had become used to the dividends. The clientele effect alsoprovides an alternative argument for the irrelevance of dividend policy, atleast when it comes to valuation. In summary, if investors migrate to firms thatpay the dividends that most closely match their needs, no firm’s value should beaffected by its dividend policy. Thus, a firm that pays no or low dividendsshould not be penalized for doing so, because its investors do not wantdividends. Conversely, a firm that pays high dividends should not have a lowervalue, since its investors like dividends. This argument assumes that there areenough investors in each dividend clientele to allow firms to be fairly valued,no matter what their dividend policy. Empirical Evidence on the Clientele EffectResearchers have investigated whether the clientele effect is strong enough toseparate the value of stocks from dividend policy. If there is a strong enoughclientele effect, the returns on stocks should not be affected, over longperiods, by the dividend payouts of the underlying firms. If there is a taxdisadvantage associated with dividends, the returns on stocks that pay highdividends should be higher than the returns on stocks that pay low dividends, tocompensate for the tax differences. Finally, if there is an overwhelmingpreference for dividends, these patterns should be reversed. In their study ofthe clientele effect, Black and Scholes (1974) created 25 portfolios of NYSEstocks, classifying firms into five quintiles based upon dividend yield, andthen subdivided each group into five additional groups based upon risk (beta) 3737 each year for 35 years, from 1931 to 1966.14 When they regressed totalreturns on these portfolios against the dividend yields, the authors found nostatistically significant relationship between the two. These findings werecontested in a later study by Litzenberger and Ramaswamy (1979), who usedupdated dividend yields every month and examined whether the total returns inex-dividend months were correlated with dividend yields.15 They found a strongpositive relationship between total returns and dividend yields, supporting thehypothesis that investors are averse to dividends. They also estimated that theimplied tax differential between capital gains and dividends was approximately23%. Miller and Scholes (1981) countered by arguing that this finding wascontaminated by the stock price effects of dividend increases and decreases.16In response, they removed from the sample all cases in which the dividends weredeclared and paid in the same month and concluded that the implied taxdifferential was only 4%, which was not significantly different from zero. Inthe interests of fairness, we should point out that most studies of theclientele effect have concluded that total returns and dividend yields arepositively correlated. Although many of them contend that this is true becausethe implied tax differential between dividends and capital gains issignificantly different from zero, there are alternative explanations for thephenomena. In particular, while one may disagree with Miller and Scholes’conclusions, their argument - that the higher returns on stocks that pay highdividends might have nothing to do with the tax disadvantages associated withdividends but may instead be a reflection of the price increases associated withunexpected dividend increases - has both a theoretical and an empirical basis,as discussed below. 10.8. ☞: Dividend Clientele and Changing Dividend PolicyPhone companies in the United States have for long had the following features -they are regulated, have stable earnings, low reinvestment needs and pay highdividends. Many of 14 Black, F. and M. Scholes, 1974, The Effects of DividendYield and Dividend Policy on Common Stock Prices and Returns, Journal ofFinancial Economics, v1, 1-22. 15 Litzenberger, R.H. and K. Ramaswamy, 1979, TheEffect of Personal Taxes and Dividends on Capital Asset Prices: Theory andEmpirical Evidence, Journal of Financial Economics, Vol 7, 163-196. 38 38 thesephone companies are now considering entering the multimedia age and becomingentertainment companies, which requires more reinvestment and creates morevolatility in earnings. If you were the CEO of the phone company, would you a.announce an immediate cut in dividends as part of a major capital investmentplan b. continue to pay high dividends, and use new stock issues to finance theexpansion c. something else: Explain. Dividends operate as a information signalFinancial markets examine every action a firm takes for implications for futurecash flows and firm value. When firms announce changes in dividend policy, theyare conveying information to markets, whether they intend to or not. Financialmarkets tend to view announcements made by firms about their future prospectswith a great deal of skepticism, since firms routinely make exaggerated claims.At the same time, some firms with good projects are under valued by markets. Howdo such firms convey information credibly to markets? Signaling theory suggeststhat these firms need to take actions that cannot be easily imitated by firmswithout good projects. Increasing dividends is viewed as one such action. Byincreasing dividends, firms create a cost to themselves, since they commit topaying these dividends in the long term. Their willingness to make thiscommitment indicates to investors that they believe they have the capacity togenerate these cash flows in the long term. This positive signal shouldtherefore lead investors to reevaluate the cash flows and firm values andincrease the stock price. Decreasing dividends is a negative signal, largelybecause firms are reluctant to cut dividends. Thus, when a firm take thisaction, markets see it as an indication that this firm is in substantial andlong-term financial trouble. Consequently, such actions lead to a drop in stockprices. 16 Miller, M. H. and M. S. Scholes, 1978, Dividends And Taxes, Journalof Financial Economics, v6(4), 333-364. 39 39 The empirical evidence concerningprice reactions to dividend increases and decreases is consistent, at least onaverage, with this signaling theory. Figure 10.15 summarizes the average excessreturns around dividend changes for firms.17 Figure 10.15: Excess Returns aroundAnnouncements of Dividend Changes Source: Aharony and Swary We should view thisexplanation for dividends increases and decreases cautiously, however. Althoughit is true that firms with good projects may use dividend increases to conveyinformation to financial markets, given the substantial tax liability thatincreased dividends create for stockholders, is it the most efficient way? Forsmaller firms, which have relatively few signals available to them, the answermight be yes. For larger firms, which have many ways of conveying information tomarkets, dividends might not be the least expensive or the most effectivesignals. For instance, the information may be more effectively and economicallyconveyed through an analyst report on the company. There is another reason forskepticism. An equally plausible story can be told about how an increase individends sends a negative signal to financial markets. Consider 17 Aharony, J.and I. Swary, 1981, Quarterly Dividends and Earnings Announcements andStockholders' 40 40 a firm that has never paid dividends in the past but hasregistered extraordinary growth and high returns on its projects. When this firmfirst starts paying dividends, its stockholders may consider this an indicationthat the firm’s projects are neither as plentiful nor as lucrative as they usedto be. However, Palepu and Healy find that the initiation of dividends does notsignal a decline in earnings growth in a study of 151 firms from 1970 to 1979.1810.9. ☞: Dividends as Signals Silicon Electronics, a company with a history ofnot paying dividends, high earnings growth and reinvestment back into thecompany, announces that it will be initiating dividends. You would expect a. thestock price to go up b. the stock price to go down c. the stock price to remainunchanged Explain. Dividend policy is a tool for changing financing mix Dividendpolicy cannot be analyzed in a vacuum. Firms can use dividend policy as a toolto change their debt ratios, In chapter 9, we examined how firms that want toincrease or decrease leverage can do so by changing their dividend policy:increasing dividends increases leverage over time, and decreasing dividendsreduces leverage. When dividends increase, stockholders sometimes get a bonus inthe form of a wealth transfer from lenders to the firm. Lenders would ratherhave firms accumulate cash than pay it out as dividends. The payment ofdividends takes cash out of the firm, and this cash could have been used tocover outstanding interest or principal payments. Not surprisingly, bond pricesdecline on the announcement of large increases in dividends. It is equityinvestors who gain from the loss in market value faced by bondholders.Bondholders, of course, try to protect themselves against this loss byrestricting how much firms can pay out in dividends. Returns: An EmpiricalAnalysis, Journal of Finance, Vol 36, 1-12. 41 41 Dividends reduce managerialdiscretion/power In examining debt policy, we noted that one reason forincreasing debt levels was to induce managers to be more disciplined in theirproject choice. Implicit in this free cash flow argument is the assumption thatcash accumulations, if left to the discretion of the managers of the firm, wouldbe wasted on poor projects. If this is true, then forcing a firm to make acommitment to pay dividends would be an alternative to forcing managers to bedisciplined in project choice and to reducing the cash that is available fordiscretionary uses. If this is the reason stockholders want managers to committo paying larger dividends, then in firms where there is a clear separationbetween ownership and management, managers should pay larger dividends than infirms with substantial insider ownership and involvement in managerialdecisions. Managerial Interests and Dividend Policy We have considered dividendpolicy in this chapter almost entirely from the perspective of equity investorsin the firm. In reality, though, it is managers who set dividend policy and itshould come as no surprise that there may be a potential for a conflict ofinterests between stockholders and managers. The Source of the Conflict Inexamining debt policy, we noted that one reason for taking on more debt was toinduce managers to be more disciplined in their project choice. Implicit in thisfree cash flow argument is the assumption that cash accumulations, if left tothe discretion of the managers of the firm, would be wasted on poor projects. Ifthis is true, we can argue that forcing a firm to make a commitment to paydividends provides an alternative to forcing managers to be disciplined inproject choice and to reducing the cash that is available for discretionaryuses. If this is the reason stockholders want managers to commit to payinglarger dividends, firms in which there is a clear separation between ownershipand management, 18 Palepu, K and P. Healy, 1988, Earnings Information Conveyedby Dividend Initiations and Omissions, Journal of Financial Economics, v21,149-175. 42 42 should pay larger dividends than should firms with substantialinsider ownership and involvement in managerial decisions. What do managersbelieve about dividend policy? Given the pros and cons for paying dividends, andthe lack of a consensus on the effect of dividends on value, it is worthconsidering what managers factor in when they make dividend decisions. Baker,Farrely and Edelman (1985) surveyed managers on their views on dividend policyand reported the level of agreement with a series of statements. Table 10.3summarizes their findings – Table 10.3: Management Beliefs about Dividend PolicyStatement of Management Beliefs Agree No Opinion Disagree 1. A firm's dividendpayout ratio affects the price of the stock. 61% 33% 6% 2. Dividend paymentsprovide a signaling device of future prospects. 52% 41% 7% 3. The market usesdivided announcements as information for assessing firm value. 43% 51% 6% 4.Investors have different perceptions of the relative riskiness of dividends andretained earnings. 56% 42% 2% 5. Investors are basically indifferent with regardto returns from dividends and capital gains. 6% 30% 64% 6. A stockholder isattracted to firms that have dividend policies appropriate to the stockholder’stax environment. 44% 49% 7% 7. Management should be responsive to shareholders'preferences regarding dividends. 41% 49% 10% It is quite clear from this surveythat, rightly or wrongly, managers believe, that their dividend payout ratiosaffect firm value and operate as signals of future prospects. They also operateunder the presumption that investors choose firms with dividend policies thatmatch their preferences and that management should be responsive to their needs.43 43 In an updated and comprehensive survey19 of dividend policy published in2004, Brav, Graham, Harvey and Michaely conclude that management’s focus is noton the level of dividends but on changes in these dividends. Indicating a shiftfrom views in prior studies, many managers in this survey saw little gain fromincreasing dividends even in response to higher earnings and preferred stockbuybacks instead. In fact, many managers in companies that paid dividendsindicated regret the level of dividends paid by their firms, indicating thatthey would have set the dividend at a much lower level, if they had the choice.In contrast to the survey quoted in the last paragraph, managers also rejectedthe idea that dividends operate as useful financial signals, From the survey,the authors conclude that the rules of the game for dividends are the following:do not cut dividends, have a dividend policy similar to your peer group,preserve a good credit rating, maintain flexibility and do not take actions thatreduce earnings per share. 10.10. ☞: Corporate Governance and Dividend Policy Incountries, where stockholders have little or no control over incumbent managers,you would expect dividends paid by companies a. to be lower than dividends paidin other countries b. to be higher than dividends paid in other countries c. tobe about the same as dividends paid in other countries Conclusion There arethree schools of thought on dividend policy. The first is that dividends areneutral, and that they neither increase nor decrease value. Stockholderstherefore are indifferent between receiving dividends and enjoying priceappreciation. This view is based upon the assumptions that there are no taxdisadvantages to investors associated with receiving dividends, relative tocapital gains, and that firms can raise external capital for new investmentswithout issuance costs. The second view is that dividends destroy value forstockholders, because they are taxed at much higher rates than capital gains.The evidence for this tax disadvantage 19 Brav, A., J.R. Graham, C.R. Harvey andR. Michaely, Payout Policy in the 21st Century, 2004,, Working Paper, DukeUniversity. 44 44 is strong both in the tax code and in markets, when we examinehow stock prices change on ex-dividend days. On average, stock prices decline byless than the amount of the dividend, suggesting that stockholders in most firmsconsider dividends to be less attractive than equivalent capital gains. Thethird school of thought makes the argument that dividends can be valueincreasing, at least for some firms. In particular, firms that have accumulatedstockholders who prefer dividends to capital gains should continue to pay largeand increasing dividends to keep their investor clientele happy. Furthermore,increasing dividends can operate as a positive signal to financial markets andallow a firm to change its financing mix over time. Finally, forcing firms topay out dividends reduces the cash available to managers for new investments. Ifmanagers are not investing with the objective of maximizing stockholder wealth,this can make stockholders better off. In summary, there is some truth to allthese viewpoints, and it may be possible to develop a consensus around thepoints on which they agree. The reality is that dividend policy requires atrade-off between the additional tax liability it may create for firms and thepotential signaling and free cash flow benefits of making the additionalcommitment to their stockholders. In some cases, the firm may choose not toincrease or initiate dividends, because its stockholders are in high taxbrackets and are particularly averse to dividends. In other cases, dividendincreases may result. 45 45 Live Case Study The Trade Off on Dividend PolicyObjective: To examine how much cash your firm has returned to its stockholdersand in what form (dividends or stock buybacks), and to evaluate whether thetrade off favors returning more or less. Key Questions: • Has this firm everpaid out dividends? If yes, is there a pattern to the dividends over time? •Given this firm’s characteristics today, do you think that this firm should bepaying more dividends, less dividends or no dividends at all? Framework forAnalysis: 1. Historical Dividend Policy • How much has this company paid individends over the last few years? • How have these dividends related toearnings in these years? 2. Firm Characteristics • How easily can the firmconvey information to financial markets? In other words, how necessary is it forthem to use dividend policy as a signal? • Who are the marginal stockholdera inthis firm? Do they like dividends or would they prefer stock buybacks? • Howwell can this firm forecast its future financing needs? How valuable ispreserving flexibility to this firm? • Are there any significant bond covenantsthat you know of that restrict the firm’s dividend policy? • How does this firmcompare with other firms in the sector in terms of dividend policy? GettingInformation on dividend policy You can get information about dividends paid backover time from the financial statements of the firm. (The statement of changesin cash flows is usually the best 46 46 source.) To find typical dividend payoutratios and yields for the sector in which this firm operates examine the dataset on industry averages on my web site. Online sources of information: http://www.stern.nyu.edu/~adamodar/cfin2E/project/data.htm47 47 Problems 1. If Consolidated Power is priced at $50.00 with dividend, andits price falls to $46.50 when a dividend of $5.00 is paid, what is the impliedmarginal rate of personal taxes for its stockholders? Assume that the tax oncapital gains is 40% of the personal income tax. 2. You are comparing thedividend policies of three dividend-paying utilities. You have collected thefollowing information on the ex-dividend behavior of these firms. NE Gas SE BellWestern Electric Price before 50 70 100 Price after 48 67 95 Dividends/share 4 45 If you were a tax-exempt investor, which company would you use to make“dividend arbitrage” profits? How would you go about doing so? 3. Southern Railhas just declared a dividend of $ 1. The average investor in Southern Rail facesan ordinary tax rate of 50%. While the capital gains rate is also 50%, it isbelieved that the investor gets the advantage of deferring this tax until futureyears (The effective capital gains rate will therefore be 50% discounted back tothe present). If the price of the stock before the ex-dividend day is $10 and itdrops to $9.20 by the end of the ex-dividend day, how many years is the averageinvestor deferring capital gains taxes? (Assume that the opportunity cost usedby the investor in evaluating future cashflows is 10%.) 4. LMN Corporation, areal estate corporation, is planning to pay a dividend of $0.50 per share. Mostof the investors in LMN corporation are other corporations that pay 40% of theirordinary income and 28% of their capital gains as taxes. However, they areallowed to exempt 85% of the dividends they receive from taxes. If the sharesare selling at $10 per share, how much would you expect the stock price to dropon the ex-dividend day? 5. UJ Gas is a utility that has followed a policy ofincreasing dividends every quarter by 5% over dividends in the prior year. Thecompany announces that it will increase 48 48 quarterly dividends from $1.00 to$ 1.02 next quarter. What price reaction would you expect to the announcement?Why? 6. Microsoft Corporation, which has had a history of high growth and paysno dividends, announces that it will start paying dividends next quarter. Howwould you expect its stock price to react to the announcement? Why? 7. JCAutomobiles is a small auto parts manufacturing firm, which has paid $1.00 inannual dividends each year for the last 5 years. It announces that dividendswill increase to $ 1.25 next year. What would you expect the price reaction tobe? Why? If your answer is different from the prior problem, explain the reasonsfor the difference. 8. Would your answer be different for the previous problem,if JC Automobiles were a large firm followed by 35 analysts? Why or why not? 9.WeeMart Corporation, a retailer of children’s clothes, announces a cut individends following a year in which both revenues and earning droppedsignificantly. How would you expect its stock price to react? Explain. 10. RJRNabisco, in response to stockholder pressure in 1996, announced a significantincrease in dividends paid to stockholders, financed by the sale of some of itsassets. What would you expect the stock price to do? Why? 11. RJR Nabsico alsohad $ 10 billion in bonds outstanding at the time of the dividend increase inproblem 10. How would you expect Nabisco’s bonds to react to the announcement?Why? 12. When firms increase dividends, stock prices tend to increase. Onereason given for this price reaction is that dividends operate as a positivesignal. What is the increase in dividends signaling to markets? Will marketsalways believe the signal? Why or why not?
