Dividend irrelevance, and the four reasons dividends still matter

CB1 · Business Finance · 10 min read

What a dividend actually does

A company holds cash. The shareholders own the company, so they own the cash. The company pays a dividend. Now the shareholders hold the cash directly and the company does not.

Nothing has been created. Something has been moved from one pocket to another, and both pockets belonged to the same person. The share must be worth less by exactly the amount that left — which is why a share price falls on the ex-dividend date, and why that fall is not bad news.

So the first thing to be clear about is that a dividend is not a return in addition to the share price. It is part of the share price, handed over in cash.

The irrelevance argument

Modigliani and Miller's dividend result is the same kind of argument as their capital structure one, and it comes to the same kind of conclusion:

In words: how much of the return arrives as cash and how much as a higher share price is a matter of packaging, not of value.

The reason is that a shareholder can repackage it themselves, and this is the mechanism to carry away rather than the formula.

A shareholder who wants cash and is not offered a dividend can sell a few shares. A shareholder who is handed a dividend they did not want can buy more shares with it. Either operation reproduces the other policy exactly, at no cost, in a frictionless market. So no shareholder is made better or worse off by the company's choice, because every shareholder can undo it. These are called home-made dividends, and the worked example does the arithmetic of both directions.

A second strand matters as much. If a company pays out cash it does not have spare, it must raise the money elsewhere — by issuing shares. The new shareholders pay fair value and take a slice of the future, so the existing shareholders gain the dividend and lose exactly that slice. And if the company keeps cash it has no good use for, it earns the market return on it, which is what the shareholders would have earned themselves. Either way the two paths meet.

Buybacks are the same transaction

A share buyback uses company cash to purchase its own shares from whoever wants to sell. In the frictionless world it is equivalent to a dividend, and comparing the two side by side is the clearest way to see why dividends cannot matter.

A dividend reduces the price of every share and gives every shareholder cash. A buyback at fair value leaves the price of each remaining share unchanged — the company is smaller but so is the number of shares, in the same proportion — and gives cash only to those who chose to sell. A shareholder who wants cash sells into the buyback; one who does not, does nothing, and ends up with a larger share of a smaller company, which is worth the same as before.

So the two differ in who receives the cash and in the arithmetic of the share price, and not at all in anybody's wealth. Note the practical consequence, since questions like it: a buyback raises earnings per share, because earnings are divided among fewer shares. That is not value creation. It is division.

The four frictions that make it matter

The irrelevance result is a baseline, not a description. Here is what breaks it, and each one is a mechanism rather than a caveat.

Tax

If dividends are taxed differently from capital gains, the packaging stops being neutral. Where a dividend is taxed more heavily, shareholders prefer the company to retain or buy back, and a company that pays dividends anyway is destroying value for its own owners. Where different shareholders face different rates — a pension fund and an individual, say — no single policy suits them all, and companies attract the shareholders whose tax position suits the policy they have. That is the clientele effect, and it explains why changing a long-standing policy upsets people.

Signalling

Managers know more about the company's prospects than outsiders do. A dividend is cash that must actually exist, so raising it is a costly and therefore credible statement that management expects to be able to keep paying it. Cutting it is read the same way in reverse, and far more loudly.

This is why observed dividends are far smoother than observed profits. Boards raise a dividend only when they believe the new level is sustainable, and resist cutting it well past the point where the cash would be better used elsewhere. The information in the change is doing the work, not the cash.

Transaction costs

Home-made dividends are free only in the model. Selling shares costs brokerage and may mean selling an awkward number; reinvesting a dividend has the same problem. So a shareholder who needs a regular income genuinely prefers a company that pays one, and the preference is rational rather than confused.

Agency

Cash sitting inside a company can be spent by managers. Paying it out removes the temptation to make an acquisition that flatters the board or to keep a weak division alive. On this view a dividend is a discipline, and the shareholders are paying for it with whatever the company could have earned on the cash.

Where marks are lost

Treating the ex-dividend price fall as a loss. The shareholder has the cash. Total wealth is unchanged, and saying so is the answer.

Claiming a buyback creates value because earnings per share rises. Fewer shares, same earnings. The ratio moves and nothing else does.

Arguing that a high dividend signals strength, full stop. The signal is in the change and in whether it is sustainable. A high dividend funded by borrowing signals the opposite.

Forgetting that irrelevance assumes the investment policy is fixed. If paying a dividend forces the company to turn down a worthwhile project, wealth really does fall — but that is the investment decision doing the damage, not the dividend.

Mixing the model with the frictions in one breath. State the irrelevance result and its assumptions, then take the assumptions away one at a time. A question that asks whether dividend policy matters wants both halves, in that order.

Questions students ask

Why do companies pay dividends at all, if it makes no difference?

Because the frictions are real and they do not all point the same way. Signalling and agency push towards paying; tax and transaction costs may push either way depending on the shareholder. What the irrelevance result tells you is that the answer must come from one of those, and never from the bare fact that shareholders like receiving cash.

Is a scrip dividend a dividend?

Not in substance. Issuing new shares instead of paying cash gives every shareholder a larger number of shares in a company whose value has not changed, so each share is worth proportionately less and nobody is better off. No cash leaves the company, which is exactly why a company short of cash offers one.

What is a residual dividend policy?

Take every project worth doing, fund it, and pay out whatever is left. It follows directly from the irrelevance result: if the packaging does not matter, let the investment decision drive everything and treat the dividend as the remainder. The difficulty is that the remainder is as lumpy as the investment programme, and signalling means a lumpy dividend is read as instability. Most companies compromise with a steady dividend and a variable buyback.

A worked example

An invented company. Every figure is made up for this example.

Shringar Tools Limited has 10,000 shares in issue, each worth ₹500, so the whole company is worth ₹50,00,000. It holds spare cash and is considering what to do with ₹2,00,000 of it.

Follow one shareholder throughout: Meera, who holds 1,000 shares, currently worth 1,000 × 500 = ₹5,00,000. She would like ₹20,000 in cash.

Route one: the company pays a dividend of ₹20 a share

₹2,00,000 leaves the company, which is ₹20 for each of the 10,000 shares. The company is now worth ₹48,00,000 across 10,000 shares:

Route two: no dividend, and Meera sells shares instead

The price stays at ₹500. To raise ₹20,000 she sells

₹5,00,000 either way, and identical in both components: ₹4,80,000 of shares and ₹20,000 of cash. She has manufactured the dividend herself and the company's policy made no difference to her at all.

Route three: the dividend is paid but Meera did not want it

She reinvests the ₹20,000 at the new price of ₹480:

The same figure for the third time. The repackaging runs in both directions, which is the whole of the irrelevance argument in three lines of arithmetic.

Route four: a buyback instead

The company spends the same ₹2,00,000 buying its own shares at the market price of ₹500:

The price has not moved. That is the signature of a buyback at fair value and it surprises people every time: the company shrank by 4% and the share count shrank by 4%, so the value of each share is untouched.

Meera now has a choice rather than a dividend:

Four different corporate actions, one wealth figure. And notice the second line: a shareholder who ignores a buyback entirely is left with 1,000 shares out of 9,600 instead of 1,000 out of 10,000 — a larger slice of a smaller company, worth precisely what the smaller slice of the larger company was worth.

What earnings per share does, and why it proves nothing

Suppose the company earns ₹30,00,000 a year.

Earnings per share has risen 4.17% and no shareholder is a rupee better off. The same ₹30,00,000 is being divided among fewer shares, which is arithmetic rather than performance. A board that presents that rise as the result of the buyback is describing a division, and a question that offers it as evidence of value creation is testing whether you will say so.

Now break it: add a dividend tax of 10%

Dividends are taxed at 10% in the shareholder's hands and capital gains, in this invented tax system, are not taxed at all.

Meera is ₹2,000 worse off under the dividend — exactly the 10% tax on the ₹20,000. The irrelevance has gone, and it went the instant the two packages stopped being taxed alike. That is the shape of every objection to the result: the model is not wrong, one of its assumptions simply is not true, and the size of the departure is the size of the assumption's failure.

Which is also why a buyback is attractive in a tax system like this invented one. It delivers cash to the shareholders who want it without imposing the dividend tax on the ones who do not.

Where this fits in CB1

This page explains one idea from CB1 Business Finance. The full CB1 syllabus, with what each part is examined on, sits on the course page, and past-paper style mock exams with full solutions are where you find out whether the idea has stuck under time pressure.

Other CB1 topics explained

• Why financial reporting is regulated, and who it is regulated for
• Share capital and loan capital: why each source costs what it does
• Leasing against buying, and the only comparison that settles it
• Profit is not cash, and why a growing company can run out of money
• Return on capital employed, and the two things that drive it
• Gearing: why borrowing magnifies a shareholder return in both directions
• Modigliani-Miller: why gearing does not create value, and what does
• The cost of capital, and why a project must beat the whole pot
• NPV against IRR, and the three times IRR gives the wrong answer
• Working capital: the days between paying out and being paid
• Valuing a company: discounted cash flow against comparable multiples
• Synergies and the takeover premium: who actually gets the gain
• Company tax in an appraisal, and the relief that reverses the answer
• Interest rate swaps: fixing a floating cost, and where the saving comes from
• Hedging currency risk: the forward contract and the money-market hedge

Stuck on a term rather than the idea? The actuarial glossary defines the notation used above. If something here is wrong or unclear, tell us — these pages get corrected.