Dividend Policy

1 Introduction

The financial manager's job is to maximise shareholder wealth. The value of a share is, in theory, the present value of the dividends the shareholder expects to receive from it — which is the basis of the valuation arithmetic in chapter 13 — so the dividend policy an entity adopts, and what shareholders read into it, is a financial strategy question and not an administrative one.

The most common source of finance for most entities is retained earnings. All the earnings belong to the shareholders, but most entities distribute only part of them and retain the rest to fund expansion.

This chapter deals with the theory of whether the dividend decision matters at all (sections 2 and 3), and then with the three ways of making a distribution that the syllabus names: cash dividends, scrip dividends and share repurchase programmes (sections 4 to 6).

2 Dividend irrelevance

YouTube videoLoad YouTube video

Modigliani and Miller argued, in 1961, that the dividend decision is irrelevant to the value of the entity, and that all that matters is the level of earnings and the investments the entity makes with them.

The argument runs as follows. Earnings determine what an entity is able to distribute; the entity then chooses how much to pay out and how much to retain and reinvest. A large dividend leaves little for growth; a smaller dividend leaves more, and therefore produces higher dividends later. Since it is the whole stream of expected future dividends that determines the share price, the shareholder is no better and no worse off either way — provided the retained money is invested at the return the shareholders require.

So the entity should concentrate on improving earnings and on investing well, and should not agonise over the split. A lower dividend means less cash now, but this is exactly compensated by the increase in the value of the shares that the extra investment produces.

2.1 Home-made dividends

The obvious objection is that a shareholder who was relying on the dividend income now has no cash. M&M's answer is that such a shareholder can create the income by selling a few shares — a home-made dividend — and will be no worse off, because the shares are worth more than they would have been had the dividend been paid.

The reverse works too: a shareholder who does not want the cash can reinvest the dividend in more shares. Under M&M's assumptions the two positions are perfect substitutes, so the entity's choice cannot matter.

2.2 The assumptions

• there are no transaction costs, so selling shares to create income costs nothing;

• there is no taxation, or at least dividend income and capital gains are taxed identically, so the form of the return does not matter to the investor;

• shareholders have perfect information — they know what the retained money is being invested in and what it will earn;

• the entity's investment policy is fixed and is not affected by the dividend decision. This is the assumption that is easiest to miss and hardest to do without: the whole argument is that the dividend is what is left over after the investment programme, not something that competes with it.

Each of those is questionable in the real world, and each failure is a reason dividend policy does matter — which is section 3.

2.3 The conclusion: the residual view

If M&M are right, the rule follows immediately: invest in every project with a positive net present value, and distribute whatever is left. The dividend is a residual. If there is nothing worth investing in, distribute everything; if there is a great deal worth investing in, distribute nothing, and pay a larger dividend later.

Note what the argument does not say. It does not say the dividend is irrelevant; it says the pattern of dividends is irrelevant. Shareholders still expect to be paid eventually, and a share on which no dividend will ever be paid is worth nothing on the dividend valuation model.

3 Why dividend policy matters in practice

Although it has become common for entities to retain a high proportion of earnings and pay low dividends, it is risky for an entity to change its dividend policy without considering the consequences. Three practical effects break the M&M assumptions.

3.1 The signalling effect

Shareholders do not have perfect information, so they read the dividend as evidence about what management knows. A cut alarms them, even when it results from increased retention rather than a fall in earnings, because their immediate assumption is that the entity is in trouble; and the more the board explains, the more some of them suspect. Their reaction is to lower their expectations, and the share price falls.

The effect runs both ways: an increase can lift expectations even where it results merely from lower retention. The consequence is that a dividend, once raised, is very hard to reduce, and boards therefore raise it only to a level they are confident of sustaining.

3.2 The clientele effect

A settled dividend policy — always distributing 20% of earnings, say, or always increasing the dividend by 5% a year — attracts the shareholders that policy suits, in terms of their need for income and their tax position. They are the entity's clientele.

Some investors buy for income and some for capital growth. An investor who needs income chooses entities with a record of high distributions; one who wants growth chooses entities that retain and expand. Change the policy and the existing clientele is no longer suited by it, sells, and the share price falls until a different clientele has replaced them.

M&M's answer — that a shareholder needing cash can simply sell some shares — is fine in theory but ignores transaction costs and the reasonable objection that shareholders expect the entity to pay them, not to force them to become sellers of their own holding.

3.3 Taxation

The basic choice is between high dividends with low capital growth and low dividends with high capital growth. Dividend income and capital gains are commonly taxed at different rates, at different times and with different allowances, so investors are not indifferent between the two. Which is preferable depends on the individual investor, which is why taxation is really a reason for the clientele effect rather than a separate argument.

4 Cash dividends

4.1 Features and criteria

Before an entity can pay a cash dividend, four things have to be true at once, and they are genuinely independent of one another:

1. Distributable profits. A distribution is made out of accumulated realised profits, in practice broadly the retained earnings reserve. An entity can pay a dividend in a year in which it made a loss, provided the reserve carried forward is large enough; and an entity can be highly profitable and still be unable to pay one, if past losses have exhausted the reserve.

2. Cash. Distributable profits are an accounting measure and cash is not. Profit that has gone into receivables, inventory or new machinery cannot be paid out.

3. No restriction. Debt covenants (chapter 5) commonly cap distributions, directly or through a gearing or interest-cover test.

4. The investment programme funded. A dividend paid at the expense of a positive-NPV project destroys more value than it distributes.

4.2 Dividend policies

5 Scrip dividends

A scrip dividend offers shareholders new shares instead of a cash dividend, each shareholder choosing which to take. Shareholders who take the shares receive no cash; shareholders who want the cash are paid as usual.

6 Share repurchase programmes

A share repurchase, or buy-back, is a distribution in which the entity buys its own shares from shareholders and cancels them. It is a way of returning cash to shareholders that is an alternative to a dividend, not a supplement to it.

6.1 How it is done

• Open-market purchase — the entity buys its own shares in the market over a period, usually within an announced maximum amount and subject to rules on price and on the timing of purchases. The most common method; flexible, and it can be stopped.

• Tender offer — the entity offers to buy a stated number of shares at a stated price, or invites shareholders to tender at prices within a range, and buys at the price that fills the offer. Faster and larger, and it treats all shareholders alike, but the premium has to be offered up front.

• Purchase from a specific shareholder — usually to buy out a single large holder. Because the other shareholders are not offered the same terms, it normally requires their approval.

6.2 Why an entity repurchases shares

• to return surplus cash for which the entity has no positive-NPV use, without creating the expectation of recurrence that a raised dividend creates;

• to give shareholders the choice: those who want cash sell, and those who do not simply end up owning a larger proportion of the entity;

• because in many regimes the return arrives as a capital gain rather than as dividend income, which suits some shareholders better — a clientele argument;

• to increase gearing deliberately, where the entity judges itself under-geared on the reasoning in chapter 4;

• to signal that the board considers the shares undervalued, which is a credible signal precisely because it is expensive to make;

• to offset the dilution caused by shares issued under employee share schemes;

• to increase earnings per share — which is a reason boards like buy-backs and a reason to be sceptical about them (section 6.3).

Against all of that: a repurchase permanently reduces the entity's cash and its financial flexibility; it can be read as an admission that management can find nothing better to do with the money; and if the shares turn out to have been overvalued, the shareholders who stayed have paid too much for the shares of those who left.

6.3 The effect on value, on the financial statements and on performance

7 Choosing between the three

All three distributions transfer the same value; none of them creates any. The choice is about form, timing, flexibility and who receives what.

• A cash dividend pays everyone, in proportion, and sets an expectation. Use it for the sustainable, recurring distribution — and set it at a level that can be maintained, because reducing it is expensive.

• A scrip alternative preserves the declared dividend per share while keeping the cash, and lets each shareholder choose. Use it when the dividend the entity wants to declare is larger than the cash it wants to part with.

• A share repurchase returns cash without setting an expectation and without paying shareholders who did not want to be paid. Use it for a one-off surplus, and be honest about whether the reason is the surplus or the effect on earnings per share.

8 Test your knowledge

Two quick checks before you move on: work through the flashcards to fix this chapter’s terms and definitions, then sit the objective questions for exam-style practice. Both mark themselves and explain the answers as you go.