Eduzan

Eduzan / 03 Corporate Finance

Corp 1 – Analysis of Dividends and Share Repurchases

Worked examples are fully visible. Check-yourself items are study aids you can reveal one at a time.

A dividend is a distribution paid to shareholders, authorised by the board of directors. Whether that authorisation also needs a shareholder vote depends on where the company is listed: across most of Europe shareholders must approve, while in the United States the board acts alone. This is the first of several places in this reading where ownership structures and local law change the mechanics without changing the economics.

Two features separate a dividend from a bond coupon. A dividend is discretionary rather than a legal obligation, and the amount that can legally be paid may be capped by statute or by debt contracts. Dividends and interest are also taxed differently in most jurisdictions, at both the corporate and the personal level, which is why taxation gets a section of its own later.

Shares that no longer carry the right to the next declared dividend are said to trade ex-dividend. The ex-dividend date is the first day on which the buyer of a share does not receive that dividend. All else held constant, the share price can be expected to fall on that date by the amount of the dividend, because the buyer is acquiring a claim on one less cash payment.

Payout policy is the broader term. It covers the set of principles guiding both cash dividends and the value of shares repurchased in a given year. Dividend policy is only the narrower question of whether, when and how much to pay in cash. Distinguishing them matters because a company can return cash by two routes, and the choice between those routes is one of the examinable decisions in this reading.

The contribution of dividends to long-run equity returns is not a rounding error. With dividends reinvested, the total compound annual return on the S&P 500 Index from the beginning of 1926 to the end of 2018 was 10.0%, against 5.9% measured on price alone. On the Nikkei 225 Index from 1950 to 2018 the equivalent figures were 11.1% with dividends reinvested and 8.0% on price alone. Roughly four percentage points of annual compound return sit in the distribution, in both markets.

Figure 1: The routes by which a company returns value to shareholders

Only the outer two columns are genuine distributions. The middle column re-slices the same equity into a different number of pieces.

Regular cash dividends

Most dividend payers distribute cash on a schedule, and the customary frequency varies by market. Geographic variations in payment frequency are summarised below.

Typical frequency of cash dividend payment by market

Payment rhythm	Markets where it prevails
Quarterly	Canada, United States
Semiannually	Australia, Japan, Saudi Arabia
Annually	Egypt, Germany, Thailand

A long record of consistent dividends is read by the market as evidence of consistent profitability, so companies work hard to maintain or increase the regular dividend and are reluctant to cut it during temporary difficulty. An unexpected increase in the regular dividend often lifts the share price, because management is using cash, rather than words, to communicate confidence in the future.

Extra or special dividends

An extra dividend, also called a special or irregular dividend, is either a payment by a company that has no regular schedule or a supplement on top of the regular payment. It is the tool of choice for a cyclical business that wants to share a good year without committing to repeat it. A company can declare a small regular dividend and top it up at the year end when operating results allow, conserving cash in downturns.

Two illustrations. In December 2018 Tencent Holdings, listed in Hong Kong, declared a special dividend of HKD250 million after its spin-off Tencent Music listed in New York. That special dividend was approximately 3.5% of Tencent annual dividends, and the company was a very low payer generally, yielding just 0.26% for 2018 against an average of 4.6% across all stocks listed on the Hong Kong Stock Exchange. In May 2018 Ingersoll-Rand (India) Ltd, listed in Mumbai, declared a special second interim dividend of Rs202 on top of its regular annual Rs6 dividend, having paid only the regular Rs6 for the two preceding decades except for a special Rs24 dividend in 2011. The 2018 payment came out of current year profits and accumulated surpluses, and reported year-on-year net profit growth at the time was 25%.

Example 1 · Worked

AfriSage Technologies (AST) is a hypothetical provider of commercial and enterprise software in the Southern African Development Community, reporting in South African rand. In November 2017 the board modified its dividend policy: the company targets an investment-grade long-term credit rating, the ordinary dividend shall be at least 35% of net income, and excess capital is returned to shareholders after the board considers cash at hand, projected cash flow, planned medium-term investment and capital market conditions.

AfriSage Technologies selected per share data

Measure	2018	2017
Shares in issue	632.5 million	632.5 million
Earnings, per share	ZAR14.23	ZAR12.65
Cash distributed, per share	ZAR7.61	ZAR10.68

1. What proportion of earnings did AST distribute in each of the two years?

Solution. Share count is identical in the two years, so the payout ratio can be computed per share as dividends per share divided by earnings per share.
2018: ZAR7.61 ÷ ZAR14.23 = 53.5%.
2017: ZAR10.68 ÷ ZAR12.65 = 84.4%.

2. Assuming the new policy took effect in 2018, calculate the minimum ordinary dividend for 2018 and the amount that can be treated as an extra dividend.

Solution. The floor is 35% of earnings: ZAR14.23 × 0.35 = ZAR4.98. Everything paid above that floor is discretionary excess capital, so the extra dividend is ZAR7.61 − ZAR4.98 = ZAR2.63.

Liquidating dividends

A dividend is described as liquidating when the company goes out of business and distributes net assets after all liabilities are settled, when it sells part of the business for cash and passes the proceeds to shareholders, or when it pays a dividend exceeding accumulated retained earnings and so impairs stated capital. In each case the payment is a return of capital rather than a distribution out of earnings.

Stock dividends

A stock dividend, also called a bonus issue or a scrip dividend, distributes additional shares instead of cash, typically 2% to 10% of the shares then outstanding. The total cost basis of a holding is unchanged, so the cost per share falls. A holder of 100 shares bought at US$10 has a total cost basis of US$1,000. After a 5% stock dividend the holding is 105 shares, still costing US$1,000 in total, so the cost per share falls to US$1,000 ÷ 105 = US$9.52.

Stock dividends look attractive from both sides: the shareholder receives extra shares without paying for them, the company parts with no cash, and the distribution is generally not taxable to shareholders. That appearance is misleading. Proportionate ownership does not move, because every holder receives the same proportionate increase. Value per share does not move either, because the larger share count reduces earnings per share and every other per share measure by exactly the offsetting amount.

Effect of a 3% stock dividend on a holder of 10% of a company worth US$20 million

Item	Before dividend	After dividend
Shares outstanding	1,000,000	1,030,000
Earnings per share	US$1.00	US$0.97 (1,000,000 ÷ 1,030,000)
Stock price	US$20.00	US$19.4175 (20 × 0.9709)
P/E	20	20
Total market value	US$20 million	US$20 million (1,030,000 × US$19.4175)
Shares owned	100,000 (10% × 1,000,000)	103,000 (10% × 1,030,000)
Ownership value	US$2,000,000 (100,000 × US$20)	US$2,000,000 (103,000 × US$19.4175)

Intermediate results are shown rounded to four decimal places; final results are based on full precision.

The price-to-earnings ratio is held constant in that table, which is a reasonable assumption because a stock dividend alters neither the asset base nor the earning power of the company. The same reasoning applies to a stock split. The fall in price is exactly offset by the rise in share count, so total market value is unchanged.

Companies that pay regular stock dividends do see practical advantages. The practice favours long-term investors, which can lower the cost of equity financing, and it increases the float, which improves liquidity and dampens price volatility. A traditional belief holds that a lower price attracts more investors, and US companies have often treated US$20 to US$80 as the optimal share price range, so a growing company can use regular stock dividends to stay inside it. In February 2019 Massmart, the second largest distributor of consumer goods in Africa, replaced its established practice of paying interim and final dividends in cash with a scrip dividend for the 2018 final dividend. Where the same dividend rate is then paid on the enlarged share count, dividend income rises, although the company could have achieved that directly by raising the cash dividend.

The decisive difference from the company perspective concerns the financial statements. A cash dividend reduces assets, because cash leaves, and reduces shareholders equity, because retained earnings fall. Holding everything else constant, measures of liquidity weaken. The cash ratio, which sets cash and short-term marketable securities against current liabilities, falls, and so does the current ratio, which sets all current assets against the same denominator. Gearing measures move the other way: debt against total shareholders equity rises, as does debt against total assets. A stock dividend touches neither assets nor total equity: retained earnings fall by the number of shares issued multiplied by the price per share, and contributed capital rises by the same amount. Neither stock dividends nor stock splits affect liquidity ratios or leverage ratios.

Why a stock dividend moves an equity account but a split does not

Both leave total shareholders equity unchanged, so the economic effect is identical. The accounting differs. A stock dividend is treated as a transfer of retained earnings into contributed capital, which is why the two accounts move in opposite directions by the same amount. A stock split changes only the number of shares and the par value attached to them, so no balance in any shareholder equity account is disturbed. This is exactly why a question asking which action increases contributed capital has one answer, the stock dividend, even though the question asking which action changes shareholder wealth has the answer neither.

Stock splits and reverse stock splits

A stock split has no economic effect on the company and does not change the total cost basis of a holding. In a two-for-one split each shareholder receives one additional share for each share held, so the count doubles, every per share figure halves, and both the P/E and the market value of equity are unchanged. Provided the payout ratio is maintained, the dividend yield (annual dividends per share divided by share price) is also unchanged. Two-for-one and three-for-one splits are the most common, but unusual ratios such as five-for-four or seven-for-three do occur.

Before and after a two-for-one stock split

Measure	Pre-split	Post-split
Shares in issue	4 million	8 million
Price per share	€40.00	€20.00 (€40 ÷ 2)
Equity market value	€160 million	€160 million (€20.00 × 8 million)
Earnings per share	€1.50	€0.75 (€1.50 ÷ 2)
Dividend per share	€0.50	€0.25 (€0.50 ÷ 2)
Payout ratio	1/3	1/3
Yield on the shares	1.25%	1.25% (€0.25 ÷ €20.00)
Price to earnings	26.7	26.7 (€20.00 ÷ €0.75)

A two-for-one split is economically the same event as a 100% stock dividend, since all per share data fall by 50%. Only the accounting treatment differs, as set out above. Splits are usually announced after a price rise, and many investors read the announcement as a positive sign, although more often it simply acknowledges that the price has climbed far enough to justify returning it to a more marketable range. Schneider Electric SA of France split two-for-one in 2011 and Whole Foods Market of the United States split two-for-one in 2013, in each case after a significant price rise, and in neither case was the split itself a meaningful predictor of subsequent price action.

Splitting has become less common in the United States. Constituents of the S&P 500 averaged 45 splits a year between 1980 and 2017, peaked at 114 splits in 1986, and have declined steadily since 2015, with only 5 splits in 2017. The explanation offered is the greater use of funds and exchange-traded funds by individual investors together with changes in market microstructure that have broken the link between transaction costs such as commissions and the number of shares traded, so the notion of a marketable price range has lost force.

A reverse stock split runs the process backwards: it raises the price and cuts the share count, again with no effect on the market value of equity or on the total cost basis of a holding. The objective is to lift the price into a higher and more marketable range. Companies execute reverse splits, as reported in Barron’s, to attract institutional investors and mutual funds that often avoid stocks trading below US$5. Companies in financial distress, or just out of it, use them most. Kitov Pharma, a drug developer based in Israel, said in December 2018 that it would consolidate its stock 1-for-20, taking the issued count down to 16 million so as to satisfy minimum price listing criteria on the Tel Aviv Stock Exchange and to allow its ADRs to start trading on the NASDAQ in January 2019.

Reverse splits were historically rare in Asia and are becoming more popular. They were not permitted in Japan under Corporation Law until 2001, but since 2007 the Tokyo Stock Exchange has actively encouraged them in pursuit of a standard trading lot of 100 shares for listed companies by 1 October 2018. Most companies complied ahead of the deadline; on that date 23 companies reduced their trading lot to 100 shares by carrying out reverse splits. Fuji Electric Co. Ltd announced in May 2018 that it would conduct a 1-for-5 reverse split on 1 October 2018 to bring the unit of investment into the range the exchange considers desirable, between ¥50,000 and ¥500,000.

Example 2 · Worked

In May 2018 Globus Maritime Ltd, a Greek dry bulk shipping company, was warned by NASDAQ that it no longer met continuing listing requirements because the stock had closed under the US$1 minimum on 30 business days in a row. The deadline for restoring compliance was the end of October 2018, and the company responded by announcing a 1 for 10 consolidation for 15 October. By 12 October, with the split still to happen, the stock stood at US$4.25.

1. If the reverse split took place at a price of US$4.25, find the expected price afterwards, assuming nothing else affects the split.

Solution. Ten old shares become one new share, so the value of the ten must be concentrated into the one. Expected price = 10 × US$4.25 = US$42.50.

2. Assess the claim that a consolidation of this kind leaves shareholders worse off.

Solution. The claim is not generally correct. Taken on its own, the reverse split leaves the market capitalisation of common equity unchanged. If the market reads the action as a good decision, for instance because the company retains the advantages of a NASDAQ listing, price and market capitalisation might rise. Other factors, such as continued limited growth in operations or a persistently small float and low turnover, could push the value down. The wealth effect comes from what the market infers, not from the arithmetic of the split.

Since the early 1960s theorists have argued about whether dividend policy, meaning the decisions about whether, when and how much to pay, should matter to shareholders. One camp holds that it is irrelevant, and that only decisions directly related to investment in working and fixed capital affect wealth. The other camp holds that it matters, either because investors genuinely value a unit of dividends above an equal amount of uncertain capital gain, or because of market imperfections. The imperfections usually named are taxes, asymmetric information between insiders and outsiders, and agency costs arising from a tendency for management to squander spare cash.

Dividend policy does not matter

Miller and Modigliani argued in a 1961 paper that under perfect capital market assumptions, meaning no taxes, no transaction costs and symmetric information, dividend policy affects neither the cost of capital nor shareholder wealth. Their argument starts from a fixed capital budget, so the company accepts every positive net present value project, and from an already optimal capital structure. The dividend decision is therefore independent of the investment and financing decisions.

The mechanism is easiest to see in an all-equity company that decides to pay out as a dividend exactly the amount its capital budget requires. To fund the projects it issues new shares equal in value to the capital budget, leaving the capital structure unchanged. The value of the newly issued shares exactly offsets the value of the dividend. A company paying out 5% of equity should therefore see its share price fall by 5%. An Australian share priced at A$20 before an A$1 per share dividend implies a new price of A$19. The shareholder has assets worth A$20 either way: A$20 if no dividend is paid, or A$19 of stock plus A$1 of cash if it is.

Under those assumptions there is no meaningful distinction between dividends and share repurchases. Both return cash. Suppose a company with limited investment opportunities has more current cash flow than its positive net present value projects require, and distributes the surplus.

Example 3 · Worked

Consider a company with 10,000 shares in issue. It has A$10,000 of free cash flow available now, and the future cash flows of the business are worth A$190,000 in present value terms.

1. Establish the share price, then show that distributing the free cash flow through a repurchase leaves it unchanged.

Solution. Share price = (A$10,000 + A$190,000) ÷ 10,000 = A$20.
Using the free cash flow to repurchase instead of paying an A$1 dividend, the company buys A$10,000 ÷ A$20 = 500 shares. The 9,500 shares remaining have a claim on the A$190,000 of future cash flows, so the price is A$190,000 ÷ 9,500 = A$20. Selling shareholders receive A$20 a share and non-selling shareholders continue to hold shares worth A$20.

2. The company pays nothing and the stock stays at A$20. A holder of 1,000 shares wants A$1,000 of cash this year. Show how a homemade dividend achieves it and state the cost.

Solution. The holder sells A$1,000 ÷ A$20 = 50 shares, reducing the position to 950 shares. Under the MM assumptions of no taxes and no transaction costs this exactly replicates a dividend. The cost is that second-period dividend income falls with the smaller holding: higher income in one period is paid for by exactly offsetting lower income later. The irrelevance claim is not that dividends do not matter to value, but that dividend policy does not, because the alternatives merely trade one dividend stream for another of equal present value.

Real world imperfections complicate the propositions. Companies and individuals both incur transaction costs. A company issuing new shares incurs flotation costs, meaning underwriting fees, legal costs, registration expenses and possible negative price effects, and estimates commonly put them between 4% and 10% of the money raised, varying with how big the company is and how big the issue is. A shareholder manufacturing a homemade dividend incurs transaction costs and, in some countries, capital gains taxes, while cash dividends attract taxes in most countries. Selling shares periodically to create an income stream is also fragile when prices are volatile, since a price decline forces the holder to sell more shares to raise the same amount.

Dividend policy matters: the bird in the hand argument

A second position holds that, even under perfect market assumptions, investors prefer a unit of dividends to a unit of potential capital gain from reinvested earnings, because dividends are seen as less risky. Graham, Dodd, Cottle and Tatham put it in 1962 as the view that the typical unit of reinvestment is worth less to the shareholder than a unit paid out. The label comes from the proverb about a bird in the hand. If a given amount of dividends really is less risky than the same amount of capital gain, a dividend payer should have a lower cost of equity than an otherwise identical non-payer, and therefore a higher share price.

Miller and Modigliani reject the argument on its own terms. Under their assumptions, paying or raising the dividend today does nothing to the risk of future cash flows. It only lowers the ex-dividend price of the share. The riskiness of the business is set by the assets and the investment programme, not by the timing of distributions.

Dividend policy matters: the tax argument

A third position starts from the observation that dividend income has traditionally been taxed more heavily than capital gains in some countries. United States rules since 2012 cap both long-term capital gains and dividends on shares held at least 60 days at a maximum of 20%. Mainland China takes the opposite starting point, imposing no capital gains tax on shares at all, but grading the charge on dividend income by holding period: 20% where the holding is under one month, 10% where it runs from one month to a year, and, from 2015 onward, 0% once the holding passes a year.

Where dividends are taxed more heavily than gains, a taxable investor should logically prefer companies that pay little and reinvest in profitable growth, since growth in earnings above the opportunity cost of funds should translate into a higher price. A company without enough growth opportunities to absorb retained earnings could distribute through repurchases instead. Taken to its limit, the argument implies a zero payout ratio. Practice intrudes: in some jurisdictions regulation requires distribution of excess earnings as dividends, or reclassifies repurchases as dividends where the repurchases appear to be a continuing substitute for dividend payments.

Where the evidence leaves the debate

The search for a link between dividend policy and value has not been conclusive, because so many variables affect value that isolating the dividend effect is difficult. Financial theory predicts that reinvestment opportunities should dominate, and the observed pattern is consistent with that prediction. Wherever they are located, small fast-growing companies pay out little or nothing, and mature companies with fewer reinvestment opportunities tend to pay dividends. For those mature companies, taxes, regulation and law, tradition, signalling, ownership structures and attempts to manage agency conflicts all appear to play some part in setting the payout ratio. The practical instruction for an analyst is modest but useful: judge whether a given dividend policy matches the reinvestment opportunities and the legal and financial environment the company actually faces.

Miller and Modigliani assumed symmetric information. In reality corporate managers hold more detailed information about the company than outside investors do. Once information is asymmetric, a dividend change can move the price simply because it conveys news.

A board can use the dividend to signal its view of prospects, and the signal carries more weight than a favourable statement because cash is committed. For a signal to work it has to be difficult or costly for a company without the same attributes to imitate. Dividend increases satisfy that condition: a company that does not expect its cash flows to rise cannot maintain an ever higher dividend for long, although in the short run it may borrow to fund one.

Empirical work broadly supports the pattern. Starting a dividend or raising one carries good news and tends to precede growth in earnings; skipping one or cutting it carries bad news and tends to precede earnings trouble. Declaring a dividend closes part of the information gap between those inside the company and those outside and can close part of the gap between market price and intrinsic value. Evidence from both developed and emerging equity markets shows an earnings and return effect after initiation announcements: earnings usually climb both in the year a dividend starts and over several subsequent years, and the news that a regular cash dividend is being started carries an excess return with it.

Two historical initiations after the 2008 global financial crisis make the point. Oracle Corporation initiated a US$0.05 quarterly dividend in May 2009. The annual US$0.20 dividend amounted to about US$1 billion, small beside operating cash flow of US$8 billion and a further US$9 billion of cash and cash-equivalent assets on the balance sheet at the end of fiscal year 2009. An analyst covering Oracle for institutional investors read the announcement as a signal that the company was positioned to ride out the downturn and gain market share. In mid-2009 Paris-based Groupe Eurotunnel announced its first ever dividend after completing a debt restructuring and receiving insurance proceeds from a fire that had closed the Channel Tunnel, and its chief executive described the moment as a turning point with a return to profitability anticipated.

A related explanation is attention rather than information. A dividend initiation or increase attracts scrutiny, and management of an undervalued company has an incentive to invite that scrutiny because it should produce an upward price adjustment. Management of an overvalued company has the opposite incentive, which is part of why the signal is credible.

Example 4 · Worked

Management of a company with poor future prospects recommends a dividend increase to the board, explaining that investors may then believe the company has positive prospects, raising share value and shareholder wealth.

1. State whether the imitation is likely to achieve the stated objective over the long term, and justify the answer.

Solution. It is not likely to work over the long term. Dividend increases are costly to mimic precisely because a company that does not expect cash flows to grow cannot sustain the higher payment. It must eventually cut the dividend, or return to the market for new equity or debt to fund it. Both routes are costly, because both are associated on average with downward revisions to the share price. The effectiveness of the signal rests on that cost.

Records of consistent increases

Many companies take pride in long unbroken records of dividend increases. Standard and Poor’s identifies constituents of the S&P 500 Index, the Europe 350 Index, the Pan Asia Index and the S&P/TSX Canadian Index that have raised their dividend for a number of consecutive years: at least 25 years for the S&P 500, at least 10 years for the Europe 350, at least 7 years for the Pan Asia Index and at least 5 years for the S&P/TSX. These companies span many industries and tend to share a set of characteristics:

They occupy either a dominant or a defensible niche position within their industry.
Their operations are global rather than confined to one market.
Their earnings swing less than those of comparable companies.
They earn comparatively high returns on their asset base.
They carry comparatively little debt, so covenant restrictions are unlikely to bind on the payout.

Cuts and omissions

Cuts and omissions are powerful and usually negative signals. For a company under financial or operating stress the declaration date is watched closely, and merely maintaining the dividend, or cutting it by less than feared, is normally taken as good news that the difficulties are transitory and manageable, unless investors suspect management of trying to mislead the market.

Management can in principle signal positively by cutting, although this is difficult in practice. Telstra, a major Australian telecoms company with a record of paying close to 90% of profits as dividends, announced a 30% cut in 2017, its first cut in more than 20 years. Management explained that the conserved funds would be reinvested in the business, that it was planning for the longer term, and that retaining financial flexibility was a priority given rising competition and competing technologies. The immediate reaction was a 12% share price decline as yield-focused investors exited. In retrospect the market read it as a positive signal, and institutional investors regarded Telstra as successfully using its cash flow to reorganise, one of the few cases in which a large Australian payer was cutting for reasons other than extreme financial pressure.

Example 5 · Worked

In November 2018 BT Group Plc, one of the largest providers of communication services and solutions, operating in over 170 countries, trimmed its interim dividend to 4.62 pence a share from 4.85 pence. Alongside that decision it disclosed that operating cash flow, net, had collapsed by 71% to £754 million, and that turnover was down 2% at £11.6 billion, with every division contributing to the fall. In the first six months of the year it reported a pretax profit increase to £1.3 billion from £1.1 billion a year earlier and a 2% increase in adjusted earnings before interest, taxes, depreciation and amortisation to £3.7 billion from £3.6 billion, as the company cut costs during a restructuring. One analyst described the dividend decrease as an unwelcome surprise but also a prudent move given the 71% decline in net cash, and noted that it should not take too much shine off a dividend yield that had previously stood at an attractive 6.4%. BT Group was also replacing its chief executive in February 2019, so future dividends would depend on decisions taken by new leadership. As the market absorbed the information, the share price rose 6.9% to 257 pence per share.

1. Reconcile a dividend cut with a share price rise of 6.9%.

Solution. Two separate pieces of information were released together. The cut, read in isolation, is a negative signal. The accompanying figures were mixed to positive: pretax profit rose from £1.1 billion to £1.3 billion and adjusted earnings rose 2% from £3.6 billion to £3.7 billion, while operating cash flow had collapsed 71% to £754 million. Against a 71% fall in operating cash flow, maintaining a dividend that already yielded 6.4% would have been the alarming choice. The cut removed the risk of a dividend financed from outside the business, and even after the reduction from 4.85 pence to 4.62 pence the yield stayed attractive. The price move of 6.9% to 257 pence reflects the net of these, not the dividend alone.

The eBay initiation of 2019 shows how ambiguous the signal can be. Technology companies have among the lowest dividend yields and below-average payout ratios, because research and development requirements are high, some subsectors such as integrated circuit manufacture are capital intensive, and business risk is considerable as discoveries change the product landscape. All of that argues for low or no dividends so internally generated funds go to new product development and capital investment. Some technology companies nevertheless mature, and legacy names that initiated dividends as growth slowed include Apple in 2012, Cisco in 2011, Oracle in 2009 and Microsoft in 2003. When eBay acted, the standard in still-expanding technology markets was to pay nothing at all, as Baidu, JD.com, Alibaba and Weibo each did.

The first eBay dividend was declared early in 2019, set at US$0.14 a share each quarter. That worked out to a 1.6% yield, sitting below the 1.9% then offered by Microsoft and the 2.9% offered by Cisco. It simultaneously increased its existing share repurchase programme to US$4 billion. Reaction was mixed. Some read it as an interest in broadening the investor base towards income buyers while refraining from unprofitable expansion. Others read it as an admission of maturity, an acknowledgement that reinvested earnings could no longer generate high returns, and therefore as diminished growth prospects. The dividend showed confidence in cash generation, but investors would have preferred internal investment to regenerate the core business.

Large listed corporations separate the professional managers who control operations from the outside investors who own the company. When agents and owners are different parties, managers have an incentive to maximise their own welfare at the expense of the company, because they own none or very little of it and therefore do not bear the full cost of their actions. At bottom this is another information problem: if outside investors could observe managers perfectly, the behaviour would be deterred.

Overinvestment and the free cash flow hypothesis

The managerial incentive of particular concern is the private benefit obtainable from investing in negative net present value projects. Such projects destroy economic value, but they may grow the company measured in sales or assets and so enlarge the span of control of the manager. The problem is acute when compensation is tied to sales or assets rather than to value creation, which is a flaw in corporate governance rather than in the project appraisal.

Paying dividends can alleviate the overinvestment problem. By distributing free cash flow to equity as dividends, the company constrains the ability of management to take on negative net present value projects. This concern is known as the free cash flow hypothesis of Jensen.

Whether the risk exposures are real has to be evaluated case by case. Microsoft accumulated increasingly large cash positions before initiating its dividend in 2003 but was not observed to squander money on unprofitable projects. Large cash balances can also buy financial flexibility, allowing a company to react quickly to changes in its environment, to seize unforeseen opportunities, or to survive periods of restricted credit. Ford Motor Company accumulated cash during profitable years in the 1990s and the Japanese automotive parts manufacturer Denso Corporation did the same in the late 2000s and 2010s. Industry and life-cycle conditions matter: holding cash and paying little makes sense for growing companies in rapidly changing industries and much less sense for large mature companies in relatively non-cyclical industries. Consistent with this, the market reaction to dividend change announcements is generally stronger for companies with greater potential for overinvestment.

Example 6 · Worked

Two dividend-paying companies A and B compete directly. Both are all-equity financed and both have recent dividend payout ratios averaging 35%. Governance at Company B is the weaker of the two. At B, although not at A, one person holds both the chief executive role and the chair of the board. Profitable investment opportunities for B have recently become fewer, although operating cash flow at both A and B is strong.

1. An investor holding both stocks would push for a larger payment at which of the two, and on what reasoning?

Solution. At Company B, because a dividend increase may mitigate potential overinvestment agency problems. Strong operating cash flow combined with a shrinking set of profitable growth opportunities is exactly the setting in which management is tempted to overinvest, and the relatively weak governance at B removes part of the restraint that would otherwise apply. Company A is not the answer: better growth prospects argue for retaining cash, not distributing it. Underinvestment is not the concern here, since the problem identified is too much cash chasing too few good projects.

The conflict with bondholders

A second agency conflict appears once the company is financed by debt as well as equity. Paying dividends reduces the cash cushion available for the fixed payments owed to bondholders. Where the intent behind a large payment is to move value from lenders to owners, the further consequence can be that profitable projects go unfunded. Whichever route is chosen, dividends and buybacks alike raise the probability that the debt defaults.

Bondholders respond through the indenture. A covenant restricting distributions to shareholders typically defines the maximum allowable amount of distributions over the life of the bond. That allowance is usually a positive function of current and past earnings and of new equity issued, and a negative function of dividends already paid since the bonds were issued. Covenants of this shape do not really restrict dividends so long as they are funded from new earnings or from new stock issues; what they prevent is a dividend financed by selling existing assets or by issuing new debt. Covenants specifying minimum EBITDA or EBIT coverage of interest charges are also frequently used, providing assurance that operating earnings contain a cushion for fixed charges, and others focus on balance sheet strength, for example by capping the ratio of debt to tangible net worth.

Example 7 · Worked

Electric utilities often have above average dividend yields. A large share of earnings goes out as dividends, and new stock is issued from time to time to pay for the heavy project pipeline that a capital-intensive business demands. Funding a dividend with freshly issued stock looks like poor practice, since issuing stock is costly. Work on a group of United States utilities has nevertheless found a rationale for it.

1. Explain how paying high dividends and then issuing equity can mitigate agency problems in a regulated utility.

Solution. A utility is typically the only supplier of electricity in its area, so it is regulated to prevent monopoly pricing. Regulators are expected to set rates covering operating expenses and providing investors a fair return, but regulators are usually elected or politically appointed and see ratepayers as voters, which pushes rates down. Utility shareholders therefore face two conflicts: the usual manager and shareholder conflict over perquisites and overinvestment, and a regulator and shareholder conflict over rate setting.

The recurring trip to the equity market is the monitoring device. The company pays out heavily and then issues new shares. If the market judges that shareholders are not receiving a fair return, whether because rates are too low or because managers are consuming too many perquisites, the price at which new equity can be sold falls until expected returns are restored. A depressed issue price means the company may not raise enough to expand its plant to meet growing demand, and in the extreme customer needs go unmet. Facing that outcome and the prospect of angry voters, regulators have an incentive to set fair rates. The equity market thus arbitrates conflicts between shareholders and both managers and regulators.

Theory leaves the question open, yet boards and managers still have to set a number. Six factors are repeatedly named by managers themselves as relevant to the choice:

The opportunity set for new investment.
How volatile earnings are expected to be.
The desire to keep financial flexibility.
The tax regime facing the company and its owners.
The cost of floating new securities.
Restrictions imposed by contract and by law.

Some, such as taxation, are not company specific. Others, such as contractual restrictions and expected earnings volatility, are. They interact, and the presence of one may strengthen or weaken another. Importantly, once information effects, agency problems and taxes are admitted, the independence between the investment, financing and dividend decisions assumed by Miller and Modigliani no longer holds.

Investment opportunities

All else equal, a company with many profitable projects pays out less than a company with few, because it has more uses for internally generated cash and because internally generated cash is generally cheaper than new equity. The industry shapes both the supply of opportunities and the speed with which the company must respond to them. A company able to delay projects without penalty can afford to pay out more than one that must act immediately. Technology companies have much lower average dividend yields than utilities, and the main explanation is the size and time horizon of profitable opportunities relative to annual operating cash flow: change is fast in technology, so internal funds provide valuable flexibility, whereas utilities face fewer opportunities and slower change, which points to higher payouts.

Expected volatility of future earnings

Survey work on managers has consistently found that most of them hold a target payout ratio based on long-run sustainable earnings, focus more on dividend changes than on dividend levels, and are reluctant to raise the dividend if the increase might soon have to be reversed. Findings in the United States, the United Kingdom and elsewhere show managers reluctant to cut, preferring to smooth. Smoothing means relating dividend increases to the long-term earnings growth rate even when short-term earnings are volatile. As earnings become less predictable, the chance rises that some future period will fail to cover a rise granted today, and companies in that position raise dividends less often and by less. Willingness to decrease dividends in response to investment opportunities varies between countries.

Financial flexibility

Companies may decline to initiate, or may reduce or omit, dividends in order to hold substantial cash. A strong cash position lets a company meet unforeseen operating needs and exploit opportunities with minimum delay, and it is particularly valuable during contractions when credit availability shrinks. Flexibility is best treated as a tactical consideration that grows in importance when access to liquidity is critical and the payout is relatively large.

Skanska AB of Sweden, one of the largest construction and development companies, is a clean illustration. On 8 February 2019 it announced a board proposal to cut the dividend by 30% to SKr6.00, to allow continued expansion of its project development business while maintaining the financial ability to deliver sustainable returns. The chief executive pointed to political and macroeconomic uncertainties likely to increase further, and to markets levelling out across many of its home geographies and sectors. The cut was expected to conserve SKr920 million a year. With roughly SKr19 billion of cash on hand at the time and operating cash flows that at least covered the previous dividend, the reduction was accurately described as precautionary. The share price fell 9% on the announcement and recovered quickly; within two months it stood 7% above its level before the announcement, indicating a favourable market response to a cut driven by operating uncertainty and the desire for flexibility.

Where flexibility is the dominant consideration, a company may prefer to distribute mainly through share repurchases rather than regular dividends. An open market repurchase programme carries no formal requirement that any shares actually be bought, and repurchases in general do not create the same expectation of continuation as a regular dividend.

Tax considerations

For an investor who pays tax, the tax regime shapes the decision, and no two regimes look alike. There are countries that reach both capital gains and dividend income, countries that reach dividends while leaving gains untouched, and single countries whose internal rules are intricate enough to require specialist advice. Because taxation is a major fiscal policy tool subject to politics, governments revisit dividend taxation frequently, using it to encourage or discourage retention or distribution, to redistribute income, or to pursue other political, social and investment goals. For a global investor foreign taxes can matter as much as domestic ones, and foreign tax credits in the home country can be decisive. France offers an illustration. A French-domiciled company must withhold on payments to overseas investors at the rate applied to corporate profit, which fell to 25% by 2022, and those investors can generally recover the amount as a credit in their own country, especially where a treaty against double taxation is in force. The three main taxation systems are treated in the next section.

Flotation costs

Two things make up flotation cost. One is the bundle of fees that issuing stock attracts, payable to auditors, lawyers, securities regulators, investment banks and others. The other is the possible downward pressure on price from putting more shares into the market. In percentage of gross proceeds terms they are proportionally higher for smaller companies, which issue fewer shares. Because flotation costs make new equity more expensive than internally generated funds, many companies avoid setting a dividend at a level that would force them to raise new equity to finance positive net present value projects.

Example 8 · Worked

Boar’s Head Spirits Ltd, based in the United Kingdom, currently pays no dividend on its common shares. Estimated operating cash flow is £500 million, the cost of capital has been calculated at 12%, and the analyst has identified modernisation and expansion projects with positive net present value requiring £800 million. Boar’s Head has an above average debt ratio for its industry and is reluctant to increase long-term debt in the next year.

1. Discuss whether Boar’s Head would be expected to initiate a dividend.

Solution. It would not be expected to initiate. The cost of the positive net present value projects exceeds estimated operating cash flow by £800 million − £500 million = £300 million. Internally generated funds are therefore already insufficient to fund the project set. Paying a dividend would mean rejecting positive net present value projects, unless both the project shortfall and the dividend were financed with debt, and the company is unwilling to add debt. The company could reasonably issue new shares to fund the unfunded projects, but it would not be expected to issue shares purely in order to pay a dividend.

Contractual and legal restrictions

Dividends are often constrained by law or contract. In some countries, such as Brazil, distribution is legally mandated with certain exceptions. In others, such as Canada and the United States, a dividend not specifically designated as liquidating may be restricted by an impairment of capital rule, under which the remaining assets, valued net as the balance sheet shows them, must be worth at least a stated amount tied to the capital of the company.

Contractual limits are usually imposed by bondholders through the indenture, requiring the company to maintain ratios such as interest coverage or the current ratio, or to satisfy conditions, before dividends may be paid. These covenants respond to the shareholder and bondholder agency problem and limit the ability of shareholders to expropriate wealth from bondholders. In the extreme, absent covenants or legal restrictions, management could liquidate the assets and pay the proceeds out as a liquidating dividend, leaving bondholders with nothing against which to settle their claims. Where preference shares are outstanding, common dividends cannot be paid until preference dividends are paid, and where preference dividends are cumulative, arrears must be cleared first.

Example 9 · Worked

In September 2018 Makinasi Appliances Company, a hypothetical global home appliances manufacturer that pays quarterly dividends, announced the first dividend cut in its history. The quarterly payment was to drop to US$0.70 a share, against US$1.60 twelve months earlier, and the full-year 2017 distribution had been US$6.50 a share. The reduction closed out a decade in which the dividend had risen 400% in total. Facing plunging global demand, with sales forecast to fall 19%, and ongoing competition in white goods, Makinasi expected a loss as high as US$32.5 million for the fiscal year ending March 2019, on an operating loss of US$46 million, against an analyst forecast loss of US$18.3 million for the same period. The company had already lost US$28.6 million in fiscal 2018, on an operating loss of US$30.4 million. Plans include cutting production-related costs by US$18 million and fixed costs by US$21 million, eliminating board member bonuses, reducing manager bonuses by 40%, cutting capital spending by 30% to US$27 million and cutting research and development by 13.5% to US$24 million. The company announced plans to raise up to US$50 million via a bond issue, and the national credit rating agency cut the bond rating from A to A−.

1. Discuss the decision in light of the six factors affecting dividend policy.

Solution. The two dominant factors are the volatility of future earnings and the preservation of financial flexibility. Paying the full dividend would have lowered liquidity ratios and forced Makinasi to raise even more external capital than the US$50 million bond issue already planned. It would probably also have produced a more severe downgrade than the move from A to A−, raising the cost of debt. Paying in full while reporting an expected loss of US$32.5 million, far worse than the US$18.3 million analysts had forecast, would have signalled that management was not serious about cutting costs and curtailing losses, which sits badly beside the announced cuts of US$18 million in production costs and US$21 million in fixed costs. Flotation costs also play a part: flotation costs on new equity are typically higher than on new debt, and a dividend above US$0.70 a share might have required new equity in addition to the US$50 million of debt.

Three main systems determine how corporate earnings distributed as dividends are taxed: double taxation, dividend imputation and split-rate taxation. Other regimes are combinations of these. The examinable skill is to compute the effective tax rate on one currency unit of corporate pretax earnings that is paid out as a dividend.

Double taxation

Under a double taxation system, corporate pretax earnings are taxed once at the corporate level and taxed again at the shareholder level if distributed to taxable shareholders. Each layer applies to what survives the previous one, which gives the compact form of the effective rate.

Effective tax rate under double taxation, where t_c is the corporate rate and t_p is the personal rate on dividends.

Double taxation of dividends at a 15% personal tax rate (per US$100)

Stage	Amount
Profit earned, before any tax	US$100
Rate applied at the company level	35%
Profit surviving the company charge	US$65
Amount distributed under a full payout	US$65
Personal charge on the distribution	US$9.75
Cash the shareholder retains	US$55.25
Combined burden on distributed profit	44.8%

The combined rate is (35 + 9.75) ÷ 100 = 0.4475, or 44.8%.

Investors clearly prefer a lower rate on dividends, but it is not obvious whether they prefer a higher or lower payout. That depends on whether long-term capital gains are taxed in their country at all, and on whether the gains rate sits above or below the dividend rate.

Dividend imputation

Under imputation, profit that leaves the company as a dividend bears tax only once, and the rate that applies is the personal rate of the recipient. Australia and New Zealand use this system. Earnings are first taxed at the corporate level. When distributed, the shareholder receives a tax credit, known as a franking credit, for the corporate taxes already paid on those earnings, so the corporate tax is imputed to the individual. Where the marginal rate of the shareholder exceeds the corporate rate, the shareholder pays the difference. Where it is lower, the shareholder can receive a credit for the difference.

Taxation of dividends under a tax imputation system (A$)

Item	Shareholder marginal rate 15%	Shareholder marginal rate 47%
Profit earned, before any tax	A$100	A$100
Company charge levied at 30%	30	30
Profit surviving the company charge	70	70
Amount distributed under a full payout	70	70
Personal assessment on the pretax figure	15	47
Franking credit for company tax paid	30	30
Net personal charge payable	(15)	17
Overall burden on the distribution	15/100 = 15%	47/100 = 47%

Read the last line carefully. The system applies the shareholder rate to corporate pretax income, using credits or additional taxes to absorb the difference between the corporate and shareholder rates. The corporate rate has dropped out of the effective rate entirely.

Split-rate taxation

A split-rate system, of greater historical than current importance, taxes distributed corporate earnings at a lower corporate rate than retained earnings. At the individual level dividends are taxed as ordinary income. Distributed earnings are therefore still taxed twice, but the low corporate rate on the distributed portion mitigates the penalty.

Taxation of dividends under a split-rate system (per €100)

Item	Amount
Profit earned, before any tax	€200
Portion kept inside the business	100
Charge of 35% on the retained portion	35
Portion set aside for distribution	100
Charge of 20% on the distributed portion	20
Cash actually paid out	80
Personal rate of the recipient	35%
Cash the shareholder retains	(1 − 0.35) × 80 = 52
Overall burden on the distribution	20% + (80 × 0.35)% = 48%

Example 10 · Worked

Use the three tables above to compute, in each system, the tax borne on one unit of corporate pretax earnings that is fully distributed.

1. Double taxation, corporate rate 35%, personal rate on dividends 15%.

Solution. Corporate tax on US$100 is US$35, leaving US$65. Shareholder tax is US$65 × 0.15 = US$9.75, leaving a net dividend of US$65 − US$9.75 = US$55.25. Total tax is US$35 + US$9.75 = US$44.75, so the effective rate is 44.75 ÷ 100 = 44.8%. The formula gives the same figure directly: 1 − (1 − 0.35)(1 − 0.15) = 1 − 0.5525 = 0.4475.

2. Imputation, corporate rate 30%, for a shareholder at 15% and a shareholder at 47%.

Solution. Corporate tax is A$30 on A$100, and the full A$70 is distributed. The shareholder is assessed on the pretax figure of A$100 at the personal rate, then credited with the A$30 already paid.
At 15%: assessed A$15, credit A$30, so tax due is A$15 − A$30 = (A$15), a refund. Effective rate = 15 ÷ 100 = 15%.
At 47%: assessed A$47, credit A$30, so tax due is A$47 − A$30 = A$17. Effective rate = 47 ÷ 100 = 47%.
The effective rate equals the personal rate in both cases, which is the defining property of imputation.

3. Split-rate, with 20% on distributed earnings, 35% on retained earnings, and a shareholder rate of 35%.

Solution. Of €200 of pretax earnings, €100 is retained and taxed at 35%, giving €35. The other €100 is allocated to dividends and taxed at the lower distributed rate of 20%, giving €20, so €80 is distributed. The shareholder pays 35% on the €80 received, keeping (1 − 0.35) × 80 = €52. Per €100 allocated to dividends, total tax is the €20 corporate charge plus €80 × 0.35 = €28 of personal tax, so the effective rate is 20 + 28 = 48%.

Preference for current income against capital gains

All else equal, the lower the tax rate of an investor on dividends relative to the rate on capital gains, the stronger the preference for dividends. Several considerations push against that simple ranking. The investor may hold high-payout shares inside a tax-exempt retirement account. Even where dividends are taxed more lightly than gains, capital gains taxes are deferred until the shares are sold, whereas dividend taxes fall in the year received even if the cash is reinvested. Australia and the United States are among the countries where a holding still owned when the investor dies receives either a valuation step-up or an exemption measured at that date. Then there is the institutional base. Endowments and pension funds hold large blocks of stock in most industrial countries, and because they normally pay no tax on either dividends or gains, the form in which the return arrives makes no difference to them.

A stable dividend policy pays regular dividends that generally do not reflect short-term volatility in earnings. It is by far the most common approach, for the reason already established: managers are extremely reluctant to cut. Companies following it set dividends from a long-term forecast of sustainable earnings and raise them only when earnings have moved to a sustainably higher level. If the long-run forecast is slow growth, dividends grow slowly, more or less independently of cyclical spikes in either direction. If sustainable earnings are not expected to grow, dividends stay level.

Gruppo Hera, an Italian multi-utility, illustrates the pattern. Its businesses span waste management, water, the distribution of gas, electricity and district heating, and the trading of energy alongside electricity generation. Between 2003 and 2018 its dividends per share followed an upward trajectory, and earnings declines during the period were accompanied by stable or increasing dividends, which matched its declared intention of a stable, growing payment regardless of what any single year produced. The consequence is that its payout ratio varied widely, between 52% and 125%, across the period. Management stated in 2019 that annual dividends per share would keep rising, moving from €0.10 towards €0.11 by 2022.

Figure 2: Stable dividend against constant payout ratio, on the same earnings path

The stable policy transmits only the trend in sustainable earnings. The constant payout policy transmits the full cycle, including the parts management believes are temporary.

Compared with a constant payout ratio policy, a stable policy involves less uncertainty for shareholders about the level of future dividends, because the constant payout policy passes short-term volatility in earnings and investment opportunities straight through to the payment.

The target payout adjustment model

A stable policy can be modelled as gradual adjustment towards a target payout ratio, where the target payout ratio is the proportion of earnings the company intends to distribute over the long term. John Lintner set out such a model in 1956 on the basis of three conclusions from his study of dividend behaviour: companies hold a target payout ratio based on long-term sustainable earnings; managers care more about dividend changes than about the dividend level; and cutting or scrapping a dividend happens only in extremity, as a final resort.

In its simplified form the model builds the expected increase out of four inputs. Those are the earnings expected for the coming year, the ratio the company is aiming at, the dividend paid last time, and a speed parameter equal to the reciprocal of the number of years allowed for the adjustment.

Expected increase in the dividend under the target payout adjustment model.

The same relationship written as the expected dividend for the coming year.

Example 11 · Worked

A company currently pays US$0.40, has a target payout ratio of 50% and an adjustment factor of 0.2, meaning the adjustment is to occur over five years. Expected earnings for the year ahead are US$1.50, up from US$1 last year.

1. Compute the expected increase in the dividend and the resulting dividend, then compare the growth rates of earnings and dividends.

Solution. Expected increase = (US$1.50 × 0.5 − US$0.40) × 0.2 = (US$0.75 − US$0.40) × 0.2 = US$0.07. The new dividend is US$0.40 + US$0.07 = US$0.47.

Earnings rose 50%, from US$1.00 to US$1.50, but the dividend rises only about 17.5%, from US$0.40 to US$0.47. The adjustment factor is doing exactly what it is designed to do: it releases only one fifth of the gap towards the target in any one year.

2. In the following year earnings fall temporarily from US$1.50 to US$1.34. What does the model imply?

Solution. The increase implied is (US$1.34 × 0.5 − US$0.47) × 0.2 = (US$0.67 − US$0.47) × 0.2 = US$0.04 a share, giving an implied new dividend of US$0.51. The dividend rises even though earnings fell, because at US$0.47 the payout is still below the 50% target and the implied new level continues to move the company towards it. If earnings fell further, or turned into a loss, the company would be reluctant to cut unless it revised down its estimate of sustainable earnings or its target ratio; it would rather hold the dividend flat until earnings justify the next increase.

Example 12 · Worked

Luna Inc. earned US$2.00 a share last year and paid a regular dividend of US$0.40 out of it. Earnings of US$2.80 are anticipated this year. Its target sits at 30% of earnings, and it spreads the adjustment over four years.

1. What dividend does the model imply for the current year, and what does that imply about the payout ratio?

Solution. The adjustment factor is 1/4 = 0.25.
Expected dividend = US$0.40 + [(US$2.80 × 0.3 − US$0.40) × (1/4)]
= US$0.40 + [(US$0.84 − US$0.40) × (1/4)]
= US$0.40 + US$0.11 = US$0.51, an increase of US$0.11.

Earnings are expected to rise 40% while the dividend rises 27.5%. Despite the adjustment, the payout ratio actually falls, from 20% (US$0.40 ÷ US$2.00) to 18.2% (US$0.51 ÷ US$2.80), because earnings are growing faster than the model lets the dividend catch up. The company would move towards its target ratio if earnings growth were slower or the adjustment period shorter, meaning a higher adjustment factor.

Under a constant dividend payout ratio policy the company applies a chosen payout ratio to current earnings to calculate the dividend. Dividends therefore fluctuate with earnings in the short term, and the policy is infrequently adopted in practice for exactly that reason.

Example 13 · Worked

Pampas Fertilizer is a hypothetical business, the largest producer of fertilizer in Argentina. Earnings swing sharply from quarter to quarter. Demand runs seasonally, summer being stronger than winter, and the cost base is set chiefly by ammonia prices, which move with the business cycle. Given that volatility, Pampas would have difficulty sustaining a steadily rising dividend, so in fiscal year 2018 it moved from a stable dividend policy to a constant payout ratio policy, which management calls a variable dividend policy. Pampas had paid cash dividends since 2003, at ARS1.50 per fiscal quarter, through the second quarter of fiscal 2018. Effective 30 November 2017 the board approved the variable policy: beginning with the third quarter of fiscal 2018, Pampas pays a quarterly dividend equal to 25% of net income for each quarter in which it reports net income. The stated purpose was to reflect operating results more accurately while allowing for the cyclicality of the industry.

Pampas Fertilizer earnings and dividends per share (fiscal years ending 31 March)

Quarter	Earnings per share (ARS)	Dividend per share (ARS)
2019:Q4	9.32	2.350
2019:Q3	4.60	1.152
2019:Q2	15.41	3.852
2019:Q1	10.53	2.636
2018:Q4	7.84	1.961
2018:Q3	18.65	4.660
2018:Q2	26.30	1.500
2018:Q1	21.22	1.500

1. Identify the fiscal quarter in which Pampas first applied the constant payout ratio policy.

Solution. The third quarter of fiscal year 2018. Up to and including 2018:Q2 the quarterly dividend was ARS1.500 regardless of quarterly earnings, which is visible in the table: EPS of 26.30 and 21.22 both produced ARS1.500. From 2018:Q3 onwards every payout ratio rounds to approximately 25%.

2. Demonstrate that the dividend for 2019:Q4 reflects the stated policy.

Solution. A quarter of the quarterly earnings figure is ARS9.32 ÷ 4 = ARS2.330. The reported payment of ARS2.350 sits a fraction above that, which is what you would expect given that earnings in the table carry only two decimal places.

Global trends in payout policy

Dividend practices differ internationally and change through time even within one market, which is consistent with companies adapting to changing investor tastes. Typically, fewer companies in a given United States stock market index have paid dividends than companies in a comparable European index. In some Asian markets, payouts have increased significantly from a lower base as the companies and the markets matured. Two broad trends stand out:

Across most developed markets the proportion of companies that pay any cash dividend has fallen over the long run. Canada, Japan, the United Kingdom, the United States and the European Union taken as a whole all show the pattern. Asia-Pacific is the exception on amounts rather than on counts: the value distributed each year there rose to three times its level between 2009 and 2019, whereas elsewhere the annual value merely doubled across the same decade.
Buyback participation has moved the other way. A rising proportion of companies has repurchased shares, starting from the early 1980s in the United States and roughly a decade later in continental Europe and the United Kingdom. Large Asian companies joined late but forcefully: from the closing years of the 2010s, buybacks in Japan and mainland China reached substantial size, against a history in which they were rare or absent.

Research on dividend behaviour globally shows that aggregate dividend amounts and payout ratios have generally increased over time even as the fraction of payers has decreased, with aggregate payments concentrated in a relatively small number of companies. There has been some reversal in the downward trend since the global financial crisis. Payers are on average larger, more profitable, hold fewer growth opportunities and spend less on research and development than non-payers.

A further finding is a negative relationship, documented internationally, between dividend initiations and increases on the one hand and enhanced corporate governance and transparency on the other, with compulsory IFRS adoption and the enforcement of fresh insider dealing legislation among the markers used. The reading is that dividends carry less information, and signal less, once governance and market transparency improve. Payout policies are correspondingly less generous in countries requiring detailed disclosure and offering strong investor protection. Reduced information asymmetry and reduced agency problems, together with the flexibility offered by repurchases, appear to explain the long-term decline in the number of dividend payers.

Cash dividends and share repurchases, Russell 1000 companies

Time period	Cash dividends (US$ billions)	Share repurchases (US$ billions)	CAGR cash dividends	CAGR repurchases
1Q2000 to 4Q2000	126	152	n/a	n/a
1Q2007 to 4Q2007	286	680	13.0%	25.0%
4Q2008 to 3Q2009	262	223	9.0%	4.0%
3Q2014 to 2Q2015	452	650	10.0%	11.0%

Cash dividends include special dividends. Compound annual growth rates are measured against the base year 2000 and are reported as published.

The table repays close reading. Between 2000 and 2007 repurchases grew at almost twice the rate of cash dividends, 25.0% against 13.0%. During the crisis of 2008 and 2009 discretionary buybacks were slashed, dropping from US$680 billion to US$223 billion, as operating cash flows shrank and some companies faced outright distress. Cash dividends were also cut, but far less, from US$286 billion to US$262 billion. That asymmetry is the entire argument for treating repurchases as the flexible instrument and dividends as the committed one.

By 2015 operating cash flows had recovered enough that total distributions, dividends plus repurchases, reached US$1,102 billion, surpassing the previous peak of US$966 billion in 2007. Repurchases increased nearly three times from their 2009 level to US$650 billion. Cash dividends reached US$452 billion, over 40% of total distributions, against slightly less than 30% of total distributions in 2007, where US$286 billion of US$966 billion is 29.6%. The higher proportion of dividends may reflect an increased investor appetite for dividend yield during the extended period of low, and in places negative, interest rates on fixed-income securities that prevailed in many developed countries after the crisis.

In a share repurchase, also called a buyback, the company itself becomes the buyer of its own stock. Corporate cash funds the purchase, which is precisely what separates it from a stock dividend or a split and what makes it a genuine alternative to a cash dividend. Shares issued and subsequently repurchased are classified as treasury shares if they may be reissued, or as cancelled shares if they will be retired. In either case they are excluded from dividends, from voting and from the computation of earnings per share.

Unlike cash dividends, growth in the use of repurchases has historically required enabling regulation. Repurchases became legal in the United Kingdom in 1981. American law never prohibited them outright, yet they stayed rare until 1982, when Securities and Exchange Commission rule 10b-18 shielded a repurchasing company from allegations of manipulating its own stock so long as the purchases stayed inside the terms of that rule. Other markets followed: Japan in 1995, Germany and Singapore in 1998, India and Norway in 1999, Denmark and Sweden in 2000. Many markets remain more restrictive than the United States, requiring shareholder approval of repurchase programmes, limiting repurchases to a certain fraction of outstanding shares, often 10%, restricting the permitted mechanisms, and imposing other conditions to protect creditors.

Two contrasts with dividends are worth fixing in mind. First, authorising an amount of repurchases does not commit the company to buying anything, whereas declaring a dividend does commit it to paying. Second, cash dividends are distributed proportionally to ownership, whereas repurchases generally are not: if buy orders are executed in the open market, cash reaches only those shareholders with concurrent sell orders.

The four methods

Buy in the open market. The most common method, with the company buying its own shares as conditions warrant. Flexibility is at its maximum here. Nothing in law compels the company to start or to finish the programme, and an announced programme can be dropped if cash is unexpectedly needed elsewhere, for an acquisition or for capital spending. European buybacks need a shareholder vote; American open market purchases do not. Manconi, Peyer and Vermaelen (2015), studying buybacks in 32 countries, found that companies have shareholder authorisation in place to let management buy back undervalued shares in the future, and concluded that the approval requirement does not compensate for poor corporate governance and instead limits the flexibility of management to time buybacks so as to create long-term value. Authorisations can last for years, and in the minds of many shareholders the announcement itself supports the price. Executed with competent timing, so that price impact stays small and perceived undervaluation is exploited, the method also keeps costs down.
Tender for a set quantity at a set price. Under a fixed price tender offer the company names both the number of shares it wants and the price it will pay, and that price normally sits above the market. Take an Australian company whose stock changes hands at A$37. It might invite existing holders to sell 5 million shares back to it at A$40. Should holders present more than 5 million shares, the usual response is to take a pro rata slice from each of them. Because the offer carries a fixed closing date, perhaps 30 days out, the whole exercise finishes quickly.
Dutch auction. This is again an offer put to existing holders, but rather than naming one price for one quantity the company publishes a band of prices it will accept. The auction uncovers the minimum price at which the company can buy the desired number of shares, and it pays that price to all qualifying bids. With the stock at A$37, the company might offer to buy 5 million shares in a range of A$38 to A$40. Each shareholder indicates the number of shares and the lowest price at which it would sell. Bids are accepted from A$38 upward until the 5 million target is filled, and the clearing level might land on A$39. Every holder who bid at or below that level, anywhere in the A$38 to A$39 band, then receives A$39. Like a fixed price tender offer, a Dutch auction can be completed in a short period.
Repurchase by direct negotiation. In some markets a company may negotiate with a major shareholder, often at a premium to the market price, to stop a large block from overhanging and depressing the price, or to prevent an activist shareholder from gaining board representation. In some notorious cases unsuccessful takeover attempts ended with the company buying back the shares of the would-be acquirer at a premium, a greenmail transaction, often to the detriment of the remaining shareholders. Private repurchases can also occur at a discount to the market price, reflecting the weaker negotiating position of a large investor with liquidity needs.

Outside the United States and Canada, almost all repurchases occur in the open market. Not every method is permissible under every local regime.

Example 14 · Worked

The board of British Columbia Industries, Inc. (BCII) is considering a 5 million common share repurchase programme. BCII holds a sizeable cash and marketable securities portfolio, and the current stock price is C$37. The chief financial officer wants the repurchases accomplished in a cost-effective manner. Some board members want speed, others stress flexibility.

1. Assess open market repurchases on cost, flexibility and speed.

Solution. Open market repurchases give the most flexibility. BCII can time purchases to periods when the market prices its stock below perceived intrinsic value, can change the amount repurchased, and can decline to execute the programme at all. They are made opportunistically, so cost matters more than speed. Because they can be conducted so as to minimise price impact and timed to exploit prices believed to be below intrinsic value, the method is also relatively cost effective. It is the slowest of the three.

2. Assess a tender at a set price.

Solution. This route is fast, but the price named normally has to carry a premium, and that pushes up what the buyback costs. The premium is not pure waste: it may itself provide a positive signal about the view management takes of the value of the stock.

3. Assess a Dutch auction, quantifying the cost advantage against a fixed price tender offer.

Solution. A Dutch auction generally allows the buyback at a lower price than a fixed price tender offer. A fixed price tender for 5 million shares at C$40 would cost 5 million × C$40 = C$200 million. A Dutch auction successful at C$38 would cost 5 million × C$38 = C$190 million, a saving of C$10 million. Dutch auctions are quick, though usually not as quick as fixed price tender offers.

Figure 3: Where a debt-financed repurchase turns from accretive to dilutive

The single crossing point is the earnings yield. Everything to its left is accretive to EPS, everything to its right is dilutive, and the point itself leaves EPS unchanged.

Example 15 · Worked

Takemiya Industries, a Japanese company, has been accumulating cash in recent years intending to expand in emerging Asian markets. Management and directors now believe that expansion is no longer practical and are considering a repurchase using surplus cash. Takemiya has 10 million shares outstanding, net income of ¥100 million and a share price of ¥120. Cash not needed for operations totals ¥240 million and is invested in Japanese government short-term securities earning close to zero interest. Advising on a programme of this scale, its investment bankers judge that open market purchases would clear at ¥20 above the quoted price, meaning ¥140 for each share.

1. Work out what happens to earnings per share if the idle cash is spent buying stock at ¥140.

Solution. Current EPS = ¥100 million ÷ 10 million shares = ¥10.00.
Net income is unchanged at ¥100 million, because the cash being spent was earning close to zero.
Shares repurchased = ¥240,000,000 ÷ ¥140 = approximately 1.7 million shares, leaving 8.3 million outstanding.
EPS after = ¥100 million ÷ 8.3 million = approximately ¥12.00, an increase of 20%.

Note the structure of the result. The increase happens because the cash withdrawn was contributing nothing to earnings. EPS would rise by more if the open market purchases were completed at the prevailing price without the ¥20 premium.

Example 16 · Worked

Selamat Plantations, Inc. plans to borrow Malaysian ringgit (MYR) 12 million and use the proceeds to repurchase shares.

Selamat Plantations data before the repurchase

Item	Value
Market price when the buyback runs	MYR60
Profit after tax	MYR6.6 million
Earnings per share, before	MYR3
Multiple of earnings	MYR60 ÷ MYR3 = 20
Yield on earnings	MYR3 ÷ MYR60 = 5%
Shares in issue	2.2 million
Buyback size, in shares	200,000

1. Calculate EPS after the repurchase, assuming the after-tax cost of borrowing is 5%.

Solution. Shares outstanding after the buyback are 2.2 million − 0.2 million = 2 million.
EPS after = [MYR6.6 million − (MYR12 million × 0.05)] ÷ 2 million
= [MYR6.6 million − MYR0.6 million] ÷ 2 million
= MYR6.0 million ÷ 2 million = MYR3.00.

EPS is unchanged, because the after-tax cost of debt of 5% is exactly equal to the earnings yield of MYR3 ÷ MYR60 = 5%.

2. Calculate EPS if the borrowing rate rises to 6% because of the increased financial risk of borrowing MYR12 million.

Solution. EPS after = [MYR6.6 million − (MYR12 million × 0.06)] ÷ 2 million
= [MYR6.6 million − MYR0.72 million] ÷ 2 million
= MYR5.88 million ÷ 2 million = MYR2.94.

Here the after-tax cost of debt of 6% exceeds the 5% earnings yield, so EPS falls.

To summarise the two cases: a repurchase may raise, lower or leave EPS unchanged, and which one occurs depends on how it is financed. With internal financing, EPS rises only if the funds used would not have earned their cost of capital had they been retained. With external financing, the effect is positive when the earnings yield exceeds the after-tax cost of financing, negative when it is below, and neutral at equality.

Why a higher EPS after a buyback does not mean shareholders are wealthier

Two arguments block the inference. First, the same surplus cash could instead have been paid out as a cash dividend, so the comparison is not against doing nothing. Treating total return informally as the sum of a dividend yield and a capital gains return, any capital gain arising from the EPS boost may simply replace the dividend yield that was given up. Second, a repurchase raises leverage, so the required rate of return on the equity rises too, and a higher earnings figure discounted at a higher rate need not produce a higher value. Finance theory is blunt on the point: changing EPS by changing the number of shares does not change shareholder wealth when total free cash flow is unchanged.

Treat a buyback as interchangeable with a cash dividend of the same size, so far as shareholder wealth is concerned, provided nothing else differs. In this context, all other things being equal is shorthand for the assumption that the taxation and the information content of the two do not differ. Establishing the baseline equivalence is what permits the more interesting analysis of what happens when taxation or information content does differ.

Share price after distributing cash C through a repurchase executed at the market price P₀.

Figure 4: Two routes, one total

The dividend splits the same total into cash plus a lower share price. The repurchase leaves it all in the share price, and concentrates it on fewer shares.

Example 18 · Worked

Rohit Chemical Industries, Inc. (RCII) has 10 million shares outstanding at a current market value of Rs200 per share. The board is considering two ways of distributing current free cash flow to equity of Rs500 million. One route is a one-off special payment of Rs500 million ÷ 10 million = Rs50 for every share. The other is buying back Rs500 million of stock. Assume dividends are received when the shares go ex-dividend, that any quantity of shares can be bought at Rs200, and that the taxation and information content of the two routes do not differ.

1. How is the wealth of a shareholder affected by the choice of method?

Solution.
Cash dividend. After the shares go ex-dividend, the holder of one share has Rs50 in cash and a share worth Rs200 − Rs50 = Rs150. Confirm the ex-dividend value from the market value of equity: [(10 million)(Rs200) − Rs500 million] ÷ 10 million = Rs1,500 million ÷ 10 million = Rs150. Total wealth from one share is Rs50 + Rs150 = Rs200.

Share repurchase. With Rs500 million RCII buys Rs500 million ÷ Rs200 = 2.5 million shares. The post-repurchase price is [(10 million)(Rs200) − Rs500 million] ÷ (10 million − 2.5 million) = Rs1,500 million ÷ 7.5 million = Rs200, unchanged. Total wealth from one share is Rs200, exactly as under the dividend.

Whether the shareholder actually sold into the repurchase is irrelevant to wealth. Selling realises Rs200 in cash; not selling leaves a share whose market value of Rs200 counts equally.

The lesson to carry forward is that the choice of distribution channel is not, by itself, a lever on wealth. The next example shows what happens when the equal-treatment assumption is broken.

Example 19 · Worked

The ordinary stock of AfriCitrus (AC) changes hands at South African rand (ZAR) 200, and 10 million shares are in issue. The board discovers that Kirk Mzazi has built a large stake, intending to steer operations in a direction it does not favour. One adviser proposes a private approach to Mzazi, offering to take ZAR500 million of stock off him at ZAR250 each, ZAR50 above where the market prices it. The board refuses, citing what such a deal would do to everyone else on the register.

1. Determine the effect of the proposed repurchase on the wealth of shareholders other than Mzazi.

Solution. At ZAR250 a share, ZAR500 million buys ZAR500 million ÷ ZAR250 = 2 million shares.
Post-repurchase price = [(10 million)(ZAR200) − ZAR500 million] ÷ (10 million − 2 million) = ZAR1,500 million ÷ 8 million = ZAR187.50.
Shareholders other than Mzazi lose ZAR200 − ZAR187.50 = ZAR12.50 for each share owned.

Compare with Example 18, where the price after the repurchase was unchanged. Total wealth is still conserved here, including that of Mzazi, but the premium transfers wealth from the remaining shareholders to the selling shareholder. Buying at the market price is wealth neutral; buying above it is not.

A widely held view among managers and investors alike is that buybacks add, on average, to shareholder value. The supporting evidence is that announcements are followed by materially positive excess returns, both in the window around the announcement itself and across the two years after it, with some studies tracking the effect out to five years. An explanation consistent with that finding is that managements tend to buy back stock when it is undervalued and to issue stock when it is overvalued.

Theory nonetheless treats a buyback and a cash dividend of the same size as interchangeable in their wealth effect, provided nothing else differs. The whole discussion of the choice therefore turns on what might not be equal. Repurchases can also be legally restricted, which removes the choice in some jurisdictions. Beyond the general aim of distributing free cash flow to equity, the additional reasons for repurchasing are:

Potential tax advantages.
Support for the price, and a signal that the board rates its own stock as a sound investment.
Extra room for management to act as it judges best.
A means of neutralising the dilution created when employee options are exercised.
A deliberate increase in gearing.

Tax. Where dividends are taxed more heavily than capital gains, repurchases have a straightforward tax advantage. Even where the two rates are equal, the option to defer the capital gains tax by simply declining to participate in the repurchase has value to many investors.

Signalling and price support. Management may regard its own shares as undervalued and therefore a good investment. Its market judgement can be as good or bad as anyone else, but corporate management does typically hold more information about operations and prospects than any outside investor or analyst, and open market repurchase, the dominant mechanism, lets it time purchases against the market price. An unexpected announcement of a meaningful programme can move the price as much as a better than expected earnings report. Within days of the worldwide equity collapse in October 1987, several well-known companies unveiled enormous buybacks, intending to arrest the fall and display confidence, and this may have been an important aspect of the recovery that followed. Some analysts take the opposite view, arguing that a repurchase programme can mean the company has no new profitable investment opportunities and is returning cash for want of anything better to do.

Flexibility. Unlike regular cash dividends, repurchase programmes appear not to create an expectation of continuation. In contrast to an announced dividend, an announced open market repurchase does not typically create an obligation to follow through, and the timing is at the discretion of managers. Repurchases also give shareholders flexibility, because participation is optional, which is not true of receiving a cash dividend.

Offsetting option dilution. Some companies repurchase to offset the dilution of earnings per share that arises when employee stock options are exercised. Whether stated or not, many try to repurchase at least as many shares as were issued on exercise, even though options are typically exercised at prices below the repurchase price.

Leverage. Repurchases can be used deliberately to modify the capital structure, since they increase leverage. Where the money comes from fresh borrowing rather than idle cash, the gearing effect is larger. Increasing EPS is also cited by managers as an objective, but it is problematic on two grounds. First, even where EPS does increase, the required rate of return is likely to increase too, reflecting higher leverage. Second, changing EPS by changing the share count does not affect shareholder wealth when total free cash flow is unchanged.

Example 20 · Worked

Deira Oasis Holdings Inc. (DOHI) has debt of United Arab Emirates dirham (AED) 30 million and a debt ratio of 30%. It plans a repurchase programme of AED7 million, which is 10% of the market value of its common shares.

Estimated impact on capital structure (AED millions)

Component	Start AED	Start %	Cash funded AED	Cash funded %	Debt funded AED	Debt funded %
Debt	30	30	30	32	37	37
Equity (at market)	70	70	63	68	63	63
Total capital	100	100	93	100	100	100

1. Assuming nothing else changes, what debt ratio results from financing the repurchase with cash on hand?

Solution. Equity falls by AED7 million to AED63 million and debt is unchanged at AED30 million, so total capital is AED93 million. The debt ratio rises to AED30 million ÷ AED93 million = 0.3226, or 32.3%, shown as 32% in the table.

2. What debt ratio results from financing the repurchase with new debt?

Solution. Debt rises to AED37 million and equity falls to AED63 million, so total capital stays at AED100 million. The debt ratio rises to AED37 million ÷ AED100 million = 0.3700, or 37.0%. Debt financing raises leverage substantially more than cash financing, 37% against 32%, because the numerator rises at the same time as the denominator falls.

3. Discuss the effect on the value of equity of financing with cash on hand, assuming net income and the P/E remain the same.

Solution. After the repurchase, equity stands at AED63 million on the arithmetic above. With the same net income spread over fewer shares, EPS rises. Applying the unchanged P/E to a higher EPS, the market value of equity would be expected to rise above AED63 million.

4. Discuss the effect on the value of equity of financing with new debt, assuming the after-tax cost of debt exceeds E/P, which remains the same.

Solution. Equity again stands at AED63 million immediately after the repurchase. With the after-tax cost of debt above the earnings yield, EPS falls. Applying the same P/E to a lower EPS, the market value of equity would be expected to fall below AED63 million.

5. Keeping the conditions from question 4, and adding that a rating downgrade looks imminent, what happens to the value of the debt?

Solution. After the repurchase, debt stands at AED37 million. With a real threat of a downgrade, spreads on DOHI debt against government treasuries would widen, so the market value of the debt would be expected to fall below AED37 million. Follow that through. If questions 4 and 5 both hold, equity and debt each fall in market value once the buyback is complete, which leaves the mix between them undetermined on the facts supplied.

Repurchases as a supplement and as a substitute

ITOCHU Corporation, one of the major Japanese trading houses, said in October 2018 that it would buy back stock before the fiscal year closed in March 2019, as part of lifting return on equity towards a medium to long-term goal of at least 13%. The figure named at that stage was up to ¥30 billion. In February 2019 it raised the target to ¥100 billion.

ITOCHU share buyback activity, October 2018 to March 2019

Period	Shares repurchased	Total value
December 2018 to January 2019	15,097,200	¥30 billion
February to March 2019	19,024,200	¥38 billion
Sum	34,121,400	¥68 billion

The average price over the first tranche was ¥1,987 and the average across both tranches was ¥1,993.

Other large Japanese names moved in the same direction. Toppan Printing, Tokyo Tatemono, Haseko, Sony and SoftBank each unveiled a substantial buyback in February 2019, all with the stated aim of raising return on equity, and all under activist pressure over payouts and governance standards.

Special payments and buybacks can both sit on top of the regular dividend. The natural occasion is a year in which cash flow jumps by an unusual amount that nobody expects to persist. Selecting one of these instruments tells the market that the payment, like the cash flow behind it, is a one-off. In that context a repurchase is effectively an alternative to a special cash dividend. Some companies initiate payouts using repurchases rather than dividends for the same reason, in the expectation that this will not be read as a fixed commitment.

Repurchases are also strongly procyclical. They increase when the economy is strong and cash is plentiful and fall during recessions when cash is short. Over the four years to the final quarter of 2008, S&P 500 constituents committed US$1.8 trillion to buying back stock. For comparison, over the same window they put US$2 trillion into capital projects and US$1 trillion into cash dividends. The crash of 2008 and 2009 then wiped the buyback line out. Major companies, particularly in the global financial sector, that had made large repurchases encountered challenges to their financial viability in 2008 and 2009, causing them to abandon repurchases and then to curtail or even eliminate dividends. The pattern was not confined to banking, as the Russell 1000 data in the previous section showed.

Example 21 · Worked

Shenzhen Medical Devices Ltd. (SMDL) is hypothetical, headquartered in Shenzhen, and becoming a leading supplier of testing equipment to biotechnology and pharmaceutical customers. Its primary markets are growing and it spends ¥100 million a year on research and development. SMDL is highly profitable with substantial positive free cash flow after funding positive net present value projects, and has made significant share repurchases over the past three years. Following the removal of tax on cash dividends from shares held more than a year in mainland China, management proposes initiating a cash dividend: an annual dividend of ¥0.40 a share to be paid during the next fiscal year, which on estimated earnings per share of ¥3.20 is a payout ratio of 12.5%. The proposal commits to an annual dividend thereafter, equal to or above that amount, aiming over time at a 25% payout measured against sustainable earnings, so that enough cash stays inside the business to pay for profitable capital projects. Director Z objects, on the ground that starting a cash payment would tell the market SMDL had stopped being a growth business, offering instead to support an annual 2% stock dividend. Director W, neutral between the two, tells Director Z that a 2% stock dividend will not affect the wealth of shareholders.

SMDL pro forma financial data assuming the cash dividend (¥ millions)

Earnings line	Value	Cash flow line	Value
Revenue	1,200	Operating cash generated	135
Profit before tax	155	Investing activities	(84)
Tax charge	35	Repayment of debt	(4)
Profit after tax	120	Buyback of shares	(32)
		Dividend as proposed	(15)
		Net movement in cash	0

SMDL ratios and five-year forecasts

Ratio	Value	Five-year forecast	Value
Current ratio	2.1	Revenue, annual growth	8%
Debt to equity, at market	0.27	Profit, annual growth	11%
Coverage of interest	10.8x	Cost of capital, projected	10%
Return on assets	10.0%
Return on equity	19.3%
Price to earnings	20x
Earnings yield	5.0%

1. Critique the statement of Director Z.

Solution. Four points argue against it. Dividend initiations and increases are on average associated with higher future earnings growth, not lower. The forecast sales growth of 8% and earnings growth of 11% are relatively high. SMDL still has considerable positive net present value projects available, evidenced by cash flow from investing activities of negative ¥84 million, which is consistent with substantial current growth opportunities. And SMDL has been repurchasing shares for three years, so investors already know that management distributes cash; initiating a dividend as a continuation of that policy is less likely to be read as an information signalling event.

2. Justify the statement of Director W.

Solution. A stock dividend has no effect on shareholder wealth. Each shareholder owns the same percentage of the company and of its earnings as before. Other things held constant, the quoted price adjusts downward for the distribution, and the extra shares in the holding make up the difference precisely.

3. Identify and explain the dividend policy that the proposed ¥0.40 a share dividend reflects.

Solution. The statement of cash flows shows the ¥0.40 annual dividend as a total of ¥15 million, which fully uses the free cash flow of SMDL after acceptance of positive net present value projects: 135 − 84 − 4 − 32 − 15 = 0. On its own that looks like a residual payout. But the proposal also commits to maintaining the annual dividend at ¥0.40 a share or greater, which a stable dividend policy implies, and it names a target ratio measured against sustainable earnings over the long run. Read together, the two features point to a stable policy operated through a target payout adjustment model. Setting the target as low as 25% of sustainable earnings leaves ample retained cash for profitable capital projects, which fits a company whose markets are still expanding.

No downturn since the 1930s had forced as many companies to reduce or suspend a dividend as the recession that started late in 2007. Halfway through 2009, dividends across the S&P 500 stood 25% below the year before. On the wider Russell 1000 measure, the third quarter of that year showed a fall of more than 8% against 2007. The experience was not confined to the United States: in 2009 dividends fell 15% in the United Kingdom and 9% in Australia. The analytical question in this section is how to form a judgement, in advance, on the likelihood that a cash dividend will be cut.

The traditional earnings-based measures

The traditional approach uses the dividend payout ratio and its inverse, the dividend coverage ratio.

The two measures are reciprocals, so a high payout ratio and a low coverage ratio say the same thing.

A higher payout ratio, or equivalently a lower coverage ratio, tends to indicate a higher risk of a cut, all else equal. The logic is mechanical: with a high payout ratio, a relatively small percentage decline in earnings is enough to leave the dividend unpayable out of earnings.

Example 22 · Worked

Mature European SA reports net income available for common stock of €100 million for FY2019 and dividends paid of €40 million.

1. Calculate the dividend payout and coverage ratios.

Solution. Payout ratio = 40 ÷ 100 = 40%. Coverage ratio = 100 ÷ 40 = 2.5x. The two are reciprocals, which is why only one of them needs to be computed in an exam and the other can be inverted.

Some generalisations follow from observed practice, and they should be confirmed for the particular market and period being analysed. Small young companies generally pay nothing, preferring to reinvest for growth; as they grow they typically initiate and their payout ratios tend to rise over time. Large mature companies often target payout ratios of 40% to 60%, so dividend coverage ratios run from about 1.7x to 2.5x, excluding extra payments, and mature companies are expected to sit in that range over a 5 to 10 year business cycle. Higher payout ratios, or lower coverage ratios, often constitute a risk factor that the dividend may be cut if earnings decline, and high payout ratios relative to peer group companies can point to the same concern. When coverage falls to 1.0, the dividend is considered to be in jeopardy unless a non-recurring event such as an employee strike or a typhoon explains a temporary decline in earnings. Qualitative pluses are awarded to companies with stable or increasing dividends and minuses to companies that have reduced a dividend in the past. Writing in 1962, Graham and his co-authors suggested that a history containing no cut may count for as much as a history containing many increases.

The cash flow based measure

Free cash flow to equity is the cash available for distribution as dividends after taking account of working and fixed capital needs. Ignoring those needs risks distributing cash at cross-purposes with wealth maximisation, so from that perspective free cash flow to equity, rather than reported net income, is the proper source of cash dividend payments.

Free cash flow to equity from cash flow from operations, fixed capital investment and net borrowing.

Payouts should also be considered in terms of repurchases as well as dividends, since both are cash distributions to shareholders. A comprehensive measure of safety therefore relates free cash flow to equity to the sum of the two.

A ratio of 1 means the company is returning all available cash to shareholders.

Reading the ratio: at 1, the company is returning all available cash. Significantly above 1, it is improving liquidity by building cash or marketable securities. Materially under 1, the arrangement cannot last. Distributions on that scale are being met by running down the stock of cash and marketable securities, which erodes liquidity, and sooner or later the company has to issue fresh equity or scale back its capital programme. Above-average financial leverage is a further fundamental risk factor, because additional debt issuance, whether to fund projects or to fund the dividend, may be restricted during business downturns.

Example 23 · Worked

Lygon Resources Ltd, a hypothetical company, mines and produces lithium. Its operating assets sit in South Africa, South America and Australia, and it exports worldwide. Dividends have been paid every year since 1995.

Lygon Resources selected financial information (A$ millions, years ending 31 December)

Item	2015	2016	2017	2018
Earnings, reported	540	458	399	341
Operating cash generated	837	824	679	628
Fixed capital investment	554	417	296	327
Borrowing, net of repayment	(120)	(39)	79	(7)
Cash dividends	121	256	277	323
Buybacks	0	105	277	0

1. Calculate, for each of 2015 to 2018, the dividend to earnings payout ratio, the earnings to dividend coverage ratio, free cash flow to equity, and the FCFE coverage of dividends plus repurchases.

Solution.
Payout. Dividing dividends by earnings, 2015 gives A$121 ÷ A$540 = 0.224, or 22.4%, and 2016 gives A$256 ÷ A$458 = 0.559, or 55.9%. The two later years work out at 69.4% and 94.7%.

Cover from earnings. Inverting each payout figure: A$540 ÷ A$121 gives 4.46x for 2015 and A$458 ÷ A$256 gives 1.79x for 2016, followed by 1.44x and then 1.06x.

Cash available to equity. Take operating cash flow, subtract fixed capital investment, then add net borrowing. For 2015: A$837 − A$554 + (A$120) gives A$163. For 2016: A$824 − A$417 + (A$39) gives A$368. The same arithmetic yields A$462 for 2017 and A$294 for 2018. Note that net borrowing is negative in three of the four years, so debt repayment is consuming cash that would otherwise be available to equity.

Cover of total distributions. Setting cash available to equity against dividends plus buybacks, 2015 gives A$163 ÷ (A$121 + 0) = 1.35x and 2016 gives A$368 ÷ (A$256 + A$105) = 1.02x. Repeating for the remaining years produces 0.83x and then 0.91x.

Lygon Resources coverage ratios

Measure	2015	2016	2017	2018
Payout, dividends over earnings	22.4%	55.9%	69.4%	94.7%
Cover, earnings over dividends (x)	4.46	1.79	1.44	1.06
Cash available to equity (A$ millions)	163	368	462	294
Cover of total distributions (x)	1.35	1.02	0.83	0.91

2. Discuss the trends in the two coverage measures.

Solution. Earnings to dividend coverage was nearly 4.5x in 2015 and declined steadily across the four years. By 2018 accounting earnings were just sufficient to pay the dividend, at 1.06x. An analyst watching this metric alone should already have suspected problems.

The FCFE coverage ratio was 1.35x in 2015, the year before the repurchase programme began. In 2016 coverage of dividends and repurchases fell to 1.02x, as lower capital expenditure was offset by higher dividends and the new repurchase programme. Despite falling capital expenditure and positive net borrowing of A$79 million, the ratio fell substantially further to 0.83x in 2017 as the company chose to increase distributions. Even after the repurchase programme finished, by 2018 free cash flow to equity had deteriorated so much that coverage of the dividend alone was still below 1.0x, at 0.91x.

3. Comment on the sustainability of the policy after 2017 and 2018.

Solution. With FCFE coverage at 0.83x in 2017, management would have realised that new discretionary repurchases were imprudent, and indeed repurchases stopped in 2018. By 2018 net income was still declining and FCFE coverage of the dividend alone remained below 1.0x, so management should probably consider cutting the dividend. The general lesson is that analysts should be particularly alert to companies supporting dividends and repurchases by reducing productive capital spending, by adding net debt, or by some combination of the two, because neither of these is a sustainable policy.

Figure 5: A coverage ratio crossing below one

Once the line sits below the marker, distributions are being funded from the balance sheet rather than from the year cash flow.

What past data cannot tell you

Historical figures, drawn from either the income statement or the cash flow statement, are an imperfect guide to safety. Shocks and events nobody modelled will defeat even a careful reading of the past. Equity and debt markets were shaken in 2008 and 2009 by the losses taken by almost all US and European banks, and those losses led to cuts and in some cases the virtual elimination of cash dividends. The 1962 claim by Graham and his co-authors, that for almost all ordinary shares nothing has mattered more to investment quality and value than the dividend record and its outlook, would not command universal agreement now. A weaker version does. As soon as a reduction or suspension starts to be priced in, that expectation drags on the valuation. This is why many analysts look for external market indicators of expected cuts.

An extremely high dividend yield, judged against the past record of the company and against forward-looking earnings, is often such a warning. The dividend yield on the shares of StarHub, a Singapore-listed telecoms company, was 9.4% just before its move from a fixed to a variable dividend in 2019. After the announced cut to a variable 80% of net profit for 2019 onwards, the shares were still projected to yield about 5.6%, high relative to the roughly 5% yields of recent years under the fixed dividend. At the time, shareholder equity value was anticipated to reach zero by 2020 if the fixed dividend continued. Investors in such cases bid the price down so that the expected total return after the expected cut remains adequate.

Madden (2008) supports the same caution. Examining yields for the 1,963 stocks in the MSCI World Index, of which 865 were classified as a High Dividend Universe, he found in the early months of the economic decline that 78.6% of the companies in that group had questionable ability to maintain their dividend payments, against 30.7% of all companies in the index. Later work agrees. Looking at S&P 500 constituents across the decade to 2015, the highest-yielding 5% of stocks supplied more than 8% of the worst-performing tenth. Being over-represented in that tail is what you would expect if the yields were high because fundamentals were deteriorating and the payments could not be maintained. In 2016 analysts became concerned that dividends at many European companies were unsustainable, because they were paying out the highest proportion of earnings as dividends in decades, a payout ratio of 60%, at a time when their earnings were declining. Some of those companies changed policy and cut dividends to fund reinvestment and balance sheet improvement.

The characteristics that flag an unsustainable dividend

Collecting the strands, the profile of a company that may not be able to sustain its cash dividend has been identified through this section as follows:

A payout ratio well above the 40% to 60% band typical of mature payers, or an earnings to dividend coverage ratio approaching 1.0x, particularly where the trend is deteriorating rather than flat.
An FCFE coverage ratio of dividends plus repurchases below 1.0x, meaning the distribution is being funded out of the balance sheet.
Distributions sustained by cutting productive capital spending, by adding net debt, or by both together.
Above-average financial leverage, since further debt issuance may be restricted precisely when it is needed.
A dividend yield far above the past record of the company and above what forward earnings support, which usually means the market has already priced in a cut.
A payout ratio far above the peer group, and a past record that already includes a reduction.
Volatile earnings, which raise the probability that any given level of dividend becomes uncovered in some future period.

Check yourself

1. A company pays a 4% stock dividend. State the effect on the current ratio, the debt to equity ratio, contributed capital and the proportionate ownership of an existing shareholder.

The current ratio and the debt to equity ratio are both unchanged, because no assets leave and total shareholders equity is unaltered. Contributed capital increases, because a stock dividend is accounted for as a transfer from retained earnings into contributed capital. Proportionate ownership is unchanged, because every shareholder receives the same proportionate increase in shares. Contrast this with a cash dividend, which lowers the current ratio and raises the debt to equity ratio.

2. In a jurisdiction with a 30% corporate tax rate and a 25% personal tax rate on dividends under a double taxation system, what is the effective tax rate on a unit of corporate pretax earnings that is fully distributed? How would the answer differ under imputation?

Under double taxation, 1 − (1 − 0.30)(1 − 0.25) = 1 − 0.525 = 47.5%. Under a dividend imputation system the shareholder is assessed on the pretax amount at the personal rate and credited with the corporate tax already paid, so the effective rate is simply the personal rate of 25%, and the shareholder receives a net credit of 5% of pretax earnings because the corporate rate exceeds the personal rate.

3. A company earning US$3.00 a share paid US$0.90 last year, expects US$3.60 this year, has a 40% target payout ratio and adjusts over three years. Compute the expected dividend and state whether the payout ratio moves towards the target.

The adjustment factor is 1/3. Expected dividend = US$0.90 + [(US$3.60 × 0.40 − US$0.90) × (1/3)] = US$0.90 + [(US$1.44 − US$0.90) × (1/3)] = US$0.90 + US$0.18 = US$1.08. The payout ratio moves from US$0.90 ÷ US$3.00 = 30.0% to US$1.08 ÷ US$3.60 = 30.0%. It has not moved towards the 40% target this year, because earnings grew 20% while the model released only one third of the gap. A shorter adjustment period, meaning a higher adjustment factor, would close the gap faster.

4. A company with an earnings yield of 7% is considering a debt-financed buyback at an after-tax borrowing rate of 9%. What happens to EPS, and does the answer tell you anything about shareholder wealth?

EPS falls, because the after-tax cost of the borrowed funds, 9%, exceeds the earnings yield of 7%, so the earnings lost to interest outweigh the reduction in the share count. It tells you very little about wealth. Changing the number of shares does not change wealth when total free cash flow is unchanged, and the higher leverage raises the required rate of return on the equity as well. The EPS calculation answers a mechanical question, not a valuation one.

5. Two companies each buy back shares at the market price. One trades at twice its book value per share, the other at half. What happens to book value per share in each case, and what would happen if a company traded exactly at book?

The company trading above book removes more book value per share than the share count reduction compensates for, so its book value per share falls. The company trading below book removes proportionally less, so its book value per share rises. A company repurchasing at exactly book value per share leaves book value per share unchanged, since equity and share count fall in the same proportion.

6. A board buys back a block of shares from an activist investor at a 20% premium to the market price. Compare the wealth effect on the selling and the remaining shareholders with the effect of an equivalent open market repurchase.

An open market repurchase executed at the market price leaves the price unchanged and is wealth neutral: sellers receive the market price in cash and non-sellers keep shares of the same value. A negotiated repurchase at a premium conserves total wealth but redistributes it. The block holder receives more than the shares are worth, and the post-repurchase price falls below the pre-announcement price, so continuing shareholders bear the premium. This is the greenmail pattern, and it is why boards facing this proposal often decline.

7. A mature company shows an earnings to dividend coverage ratio of 2.0x but an FCFE coverage of dividends plus repurchases of 0.75x. Which measure should drive the judgement, and what should the analyst examine next?

The cash flow measure should drive it. Earnings coverage of 2.0x sits comfortably inside the typical mature company band, but distributions are consuming more cash than the business generates for equity holders, so the shortfall is being met by drawing down cash and marketable securities. That is not sustainable, and the company will eventually have to raise equity or cut capital spending. Next the analyst should decompose free cash flow to equity to see whether it is being propped up by cuts to productive capital spending or by net new borrowing, since distributions supported by either route are the classic precursor of a cut.

8. A cyclical manufacturer wants to return an unusually large windfall to shareholders without creating an expectation that it will recur. Compare a special dividend with an open market repurchase, and note which method would be least suitable if the priority were speed.

Both signal that the distribution is one-off. A special dividend is proportionate, reaching every shareholder, and is unambiguous about being irregular. An open market repurchase carries no obligation to complete, gives management discretion over timing, lets participation remain optional for shareholders, and increases leverage slightly by reducing equity. If speed were the priority, the open market route would be the least suitable of the repurchase methods, since it works gradually; a fixed price tender offer or a Dutch auction can be completed in about 30 days, with the Dutch auction generally clearing at a lower price than a fixed price offer.

End of lesson.

Back to 03 Corporate Finance

Corp 1 – Analysis of Dividends and Share Repurchases

Overview

Forms of dividends and their effect on wealth and financial ratios

Regular cash dividends

Extra or special dividends

Liquidating dividends

Stock dividends

Stock splits and reverse stock splits

Signalling and the information content of dividend actions

Records of consistent increases

Cuts and omissions

Agency costs and their effect on payout policy

Overinvestment and the free cash flow hypothesis

The conflict with bondholders

Factors affecting dividend policy in practice

Investment opportunities

Expected volatility of future earnings

Financial flexibility

Tax considerations

Flotation costs

Contractual and legal restrictions

Taxation systems and the effective tax rate on dividends

Double taxation

Dividend imputation

Split-rate taxation

Preference for current income against capital gains

Stable dividend policy and the target payout adjustment model

The target payout adjustment model

Constant payout ratio policy and global trends in payout

Global trends in payout policy

Choosing between paying cash dividends and repurchasing shares

Repurchases as a supplement and as a substitute

Analysis of dividend safety

The traditional earnings-based measures

The cash flow based measure

What past data cannot tell you

The characteristics that flag an unsustainable dividend

Corp 1 – Analysis of Dividends and Share Repurchases

Overview

Forms of dividends and their effect on wealth and financial ratios

Regular cash dividends

Extra or special dividends

Liquidating dividends

Stock dividends

Stock splits and reverse stock splits

Theories of dividend policy and share value

Dividend policy does not matter

Dividend policy matters: the bird in the hand argument

Dividend policy matters: the tax argument

Where the evidence leaves the debate

Signalling and the information content of dividend actions

Records of consistent increases

Cuts and omissions

Agency costs and their effect on payout policy

Overinvestment and the free cash flow hypothesis

The conflict with bondholders

Factors affecting dividend policy in practice

Investment opportunities

Expected volatility of future earnings

Financial flexibility

Tax considerations

Flotation costs

Contractual and legal restrictions

Taxation systems and the effective tax rate on dividends

Double taxation

Dividend imputation

Split-rate taxation

Preference for current income against capital gains

Stable dividend policy and the target payout adjustment model

The target payout adjustment model

Constant payout ratio policy and global trends in payout

Global trends in payout policy

Share repurchase methods

The four methods

Share repurchases and earnings per share

Share repurchases and book value per share

Valuation equivalence of cash dividends and share repurchases

Choosing between paying cash dividends and repurchasing shares

Repurchases as a supplement and as a substitute

Analysis of dividend safety

The traditional earnings-based measures

The cash flow based measure

What past data cannot tell you

The characteristics that flag an unsustainable dividend