Under conditions of certainty, the Gordon growth model (P0 = Dt/ Ke - g) reveals why movements in share price relate to the profitability of a company's investment policy (business risk) and not variations in dividend policy (financial risk).
In a world of uncertainty, Gordon then explains why movements in share price relate to corporate dividend policy. Rational, risk-averse investors prefer their returns in the form of dividends now, rather than later (a "bird in the hand" philosophy).
The purpose of this Chapter is to evaluate an alternative hypothesis developed by the joint Nobel Prize winning economists, Franco Modigliani and Merton H. Miller (MM henceforth). Since 1958, their views on the irrelevance of dividend policy when valuing shares based on the economic "law of one price" have defined the development of modern finance.
The MM Dividend Irrelevancy Hypothesis
MM (1961 onwards) criticize the Gordon growth model under conditions of uncertainty supported by a wealth of empiricism, most recently the consultancy work of Stern-Stewart referenced by the author in Strategic Financial Management (op cit). According to MM, dividend policy is not a determinant of share price in reasonably efficient markets because dividends and retentions are perfect economic substitutes.
If shareholders forego a current dividend to benefit from a future retention-financed capital gain, they can still create their own home made dividends to match their consumption preferences by the sale of shares or personal borrowing and be no worse off.
If a company chooses to make a dividend distribution, it too, can still meet its investment requirements by a new issue of equity, rather than use retained earnings. So, the effect on shareholders' wealth is also neutral.
Consequently, business risk, rather than financial risk, defines all investors and management need to know about corporate economic performance.
Theoretically and mathematically, MM have no problem with Gordon under conditions of certainty. Their equity capitalisation rate (Ke) conforms to the company's class of business risk. So, as Fisher predicts (1930) share price is a function of variations in profitable corporate investment and not dividend policy. But where MM depart company from Gordon is under conditions of uncertainty.
As we explained in Chapter Three, Gordon confuses dividend policy with investment policy. For example, an increase in the dividend payout ratio, without any additional finance, reduces a firm's operating capability and vice versa. MM also assert that because uncertainty is non-quantifiable, it is logically impossible to capitalize a multi-period future stream of dividends, where K < Ke2 < Ke3 ...etc. according to the investors' perception of the unknown.
MM therefore define a current ex-div share price using the following one period model, where Ke equals the shareholders' desired rate of return (capitalisation rate) relative to the "quality" of a company's periodic earnings (class of business risk). The greater their variability, the higher the risk, the higher Ke , the lower the price and vice versa.
MM then proceed to prove that for a given investment policy of equivalent business risk (where Ke remains constant) a change in dividend policy cannot alter current share price (P0) because:
- The next ex-div price (P1) only increases by any corresponding reduction in dividend (D1) and vice versa.
To illustrate MM's dividend irrelevancy hypothesis, let us reinterpret the stock exchange data for Jovi plc, initially applied to Gordon's growth model in Chapter Three.
- With an EPS of 10 pence a full dividend distribution policy and yield of 2.5 per cent, establish Jovi's current ex-div share price using Equation (18).
- Now recalculate this price, with the same EPS forecast of 10 pence, assuming that Jovi revises its dividend policy to reinvest 100 percent of earnings in future projects with rates of return that equal its current yield.
With a policy of full dividend distribution, MM would define:
Refer back to Chapter Three and you will discover that this ex-div price is identical to that established using the Gordon growth model.
Turning to a policy of nil distribution (maximum retention) where profits are reinvested in projects of equivalent business risk (i.e. 2.5 per cent):
According to MM, because the managerial cut-off rate for investment still equals Ke, the ex-div price rise matches the fall in dividend exactly, leaving P0 unchanged.
You might care to confirm that using the Gordon growth model from the previous Chapter:
In other words, if a company does not pay a dividend, which is not unusual (particularly for high-tech growth firms), it is not possible to determine a share price.