Dividend policy in practice

Whatever a theory recommends, a real dividend decision is made inside a set of constraints. Some are legal, some are financial, and some are about the particular shareholders a company happens to have.

The legal and the liquidity constraints are worth separating carefully, because they fail in opposite directions. A company can be legally entitled to pay a dividend it cannot afford in cash — profitable on paper, with the money tied up in stock and receivables. And a company can be sitting on cash it is not permitted to distribute, because accumulated losses have wiped out its distributable reserves.

Inflation deserves attention in a Zambian context. A company earning ZMW20 million and distributing most of it may be quietly shrinking, because the machine it depreciated at 2019 prices will cost far more than that to replace. The profit was real; the capacity to keep trading was overstated.

Dividend capacity is the company’s ability to pay a dividend at all, and it has two definitions that answer two different questions.

The practical measure is the one from Lesson 1. Free cash flow to equity is what the business generated after paying its operating costs, its tax, the investment it needs to keep trading, and everything owed to its lenders. What is left is the shareholders’ cash, and it is the honest ceiling on a dividend.

A cover of 2 means the company generated twice the cash it handed out. Below 1 means it distributed more than it earned in cash, and funded the difference from reserves or borrowing.

That measure is more revealing than dividend cover on earnings, because earnings never had the capital expenditure taken out of them. A company can show earnings cover of 2.5 and cash cover of 0.8 in the same year — comfortable on the income statement, distributing money it did not generate.

Doing that once is a decision. Doing it repeatedly is a company running down its balance sheet to maintain a payout it can no longer support, which is one of the reliable early signs of trouble.

Four recognisable dividend policies, each suiting a different kind of company and each with a cost.

Stable dividend

A constant, or constantly growing, dividend each year. It gives investors a predictable cash flow, limits management’s scope to divert funds into unprofitable activities, and works well for a mature business with settled cash flows. The risk is that a bad year forces a cut, with all the signalling damage that carries.

Constant payout ratio

A fixed proportion of earnings paid out every year. It keeps a visible link between what the company earns, what it reinvests and what it distributes. But the cash flow to the investor becomes unpredictable, and it conveys nothing about what management expects — the dividend simply follows the earnings wherever they go.

Residual

A dividend only where no further positive-NPV projects remain. Popular with growing companies and with firms that have no easy access to other funds. The cost is the same as the constant payout ratio’s and worse: the cash flow is unpredictable, and the signals change every year.

Zero dividend

All surplus earnings reinvested. Common during a growth phase, and it should be reflected in a rising share price rather than in a payment. The policy has a natural end: when the growth opportunities run out, cash begins to accumulate and a new distribution policy is needed. Sitting on the cash instead is how a company becomes a takeover target.

Cash is not the only way to return value, and the alternatives solve different problems.

Share repurchase

The company uses cash to buy back its own shares. It suits a company with surplus cash and no projects to spend it on — and unlike a raised dividend it makes no promise about next year, so it does not create an expectation the company may not be able to meet.

Fewer shares in issue also means earnings are divided among fewer of them, so earnings per share rises. That is arithmetic rather than performance, in the same way the boot-strapping effect was.

Scrip dividend

Shareholders are offered new shares instead of cash, out of the current year’s earnings. Those who want income can still take the cash; those who do not take shares, and the company keeps their portion. It preserves cash without the signal a cut would send.

Scrip or bonus issue

Extra shares issued free to existing shareholders, in proportion to what they already hold, out of reserves. No wealth changes hands at all: every shareholder owns the same fraction of the same company, now divided into more pieces, so the share price falls proportionately.

The distinction between the last two matters. A scrip dividend is a real choice between cash and shares and does preserve cash. A bonus issue distributes nothing — it is a presentational act that lowers the share price into a more marketable range, and calling it an alternative to a dividend in an exam answer needs that qualification attached.

A question asking what dividend a company should pay is not asking which theory is correct. It is asking for a recommendation that survives being read by the board, and the way to build one is in a fixed order.

• Start with the investment programme. Every positive-NPV project should be funded, because that is where value comes from.
• Establish the capacity. Are there distributable profits, and is there free cash flow to equity to back the payment?
• Look at the shareholders. Who are they, what have they been led to expect, and what do they hold the shares for?
• Consider the signal. What will the market read into this figure, particularly if it is lower than last year’s?
• Then set a level the company is confident it can sustain, and say what would have to change before it moved.

The order matters. A recommendation that sets the dividend first and then finds the investment programme underfunded has the whole thing backwards — and a recommendation that funds everything and lets the dividend fall to nil has ignored what the market will make of it.

Two situations recur and are worth having answers ready for. A company that must cut should cut once, deeply enough that it will not have to cut again, and explain why — a series of small cuts is read as a company losing control. And a company enjoying an exceptionally good year should pay a special dividend rather than raise the ordinary one, so the increase is not mistaken for a new permanent level.

Underneath all of it sits the point irrelevance theory was making. A dividend policy cannot rescue a company that invests badly, and it cannot add much to one that invests well. What it can do is avoid destroying value through poor signalling and unnecessary surprises — which is a modest objective, and the right one.