The Cost of Equity, The C.A.P.M. and Management
The Cost of Equity, The C.A.P.M. and Management
The Cost of Equity, The C.A.P.M. and Management
AND MANAGEMENT
OBJECTIVES UNDER UNCERTAINTY
JOHN R. GRINYER"
INTRODUCTION
This paper considers the choice of a cost of equity capital rate, for use in capital
budgeting, under the uncertainty typically prevailing in the stock market. It
suggests that the maximisation of shareholders' wealth is likely to be achieved
more by luck than judgement, given that the shareholders' opportunity cost
probably differs from the market's required rate of return and is unknown. The
author therefore contends that the shareholders' wealth maximisation objective
is not operationally sensible as a proxy for shareholder satisficing in the multiple-
objective environment of modern business. He suggests that a shareholder
satisficing objective be explicitly specified for capital budgeting decisions, and he
considers how the Capital Asset Pricing Model (C.A.P.M.) could then be used
to estimate the rate of return required from the investment of equity funds,
using a multi-period analysis recognising the term structure of interest rates.
*The author wishes to acknowledge the helpful comments received from prof:
R.A, Brealey, h o t M. Bromwich and Mr. F. Fishwick, on early drafts of this
paper. Errors of omission andcommission are, however, his. The author is
Professor of Accountancy at Dundee L'niversity. (Paper received July 1975,
revised August 19 76)
Consider a world in which there were no transaction costs and there was certainty
concerning the cash flows associated with all investment opportunities. Given the
relevant assumptions outlined earlier, the utility of shareholders’ consumption
would be maximised if management accepted all capital projects yielding returns
greater than or equal to the single riskless rate of interest which would prevail.
Such a policy would also maximise current share price, which would be the
present value of the cash flows to be derived from owning the share, when
discounted at the riskless rate. When transactions costs are included in the
analysis, the criterion rate is found to vary depending on whether the ahareholder
is a lender (investor) or borrower (seller of shares).? It is ansumed, throughout
this paper, that he is a lender for the entire period of the project’, lifetime: m
the appropriate rate is found by reference to his alternative invmtment
opportunity, and we maximise the utility of the shareholder’s consumption by
102 John R . Grinyer
maximising his terminal wealth. Figure 1 illustrates the position assuming a two
period world and initial funds of (1-0). Curve 111 shows the cash available to the
shareholder, in periods 0 and 1, with different amounts of investment in the
physical assets available to the fm.Theslope of the curve at any point reflects
the rate of return on the marginal project at that point. Instead of investing in
physical assets the shareholder could invest in monetary claims, e.g. shares, and
the lines LLl and IL, show possible combinations of consumption available by
investing (L-0) and (1-0) respectively in this way. The slope of these lines
refleets the market’s risk free rate of return after adjusting for transaction costs.
Adoption of the maximisation of terminal wealth objective implies that the fm
should enable the shareholder to reach the highest position possible on the y axis,
and that can be done by investing (I-F) in physical assets and the balance (F-0)
in monetary assets - enabling the owner to reach point L1 in period 1. Thus, in
these circumstances, the cut off rate for physical investment should be the
market’s required risk-free rate of return, for the shareholder has always the
opportunity to invest to earn that rate elsewhere.
FIGURE 1
Y
Amount available
in Period 1.
Ll
in
Period 0
We now introduce uncertainty, but retain the assumption that the market
‘correctly’ prices shares. The opportunity cost of investment in the fm’s projects
must now be represented by the rate of return expected by the market on
alternative equities of equivalent risk to the projects, since, by assumption, such
returns equal the actual returns to be earned on those equities? so that that rate
must be the required project cut-off rate. A model which enables the calculation
of the rate of return expected by the market therefore becomes necessary as soon
as we introduce uncertainty into the analysis. Such a model is to be found in the
where
a weighting relating the premium for systematic risk on share j during period t
to the systematic risk premium for the market as a whole.
R, is the expected return from investing in the market as a whole for period 1
Rjt is the expected return on share j during period t.
it is the “risk free” rate for period t, i.e. the rate available on single-period
risk-free lending.
(R, - it) is the market premium for systematic risk, which is the risk which
cannot be diversified away.
Empirical work6 has suggested that the simple CAPM outlined above is not a
complete description of the way in which the market values capital assets. It is
also recognised that the model is based on somewhat unrealistic assumptions.’
Despite these shortcomings, the CAPM is probably the most developed
empirically-tested theory of security valuation which we currently have, and it
has gained wide academic acceptance and offers a practical approach to the
estimation of required discount rates. On these grounds it seems acceptable, as a
model of market behaviour, for use in estimating rates of return required by
investors. Clearly, ‘correctly’ priced shares are unlikely generally to exist, so we
now remove that assumption.
EXAMPLE
At a point in time, the equity market operates with a required (expected) rate of
return (R,) of lo%, and the risk free rate of interest (i) is 6%. The risk premium
for investing in the varket iis therefore 4%, and it is that rate which is required by
investors as a whole to compensate them for the disutility of risk-taking. For
ease of understanding, we will assume that cash flows derived from shareholdings
are in the form of perpetuities, so, for every €100 invested in the market,
investors expect to receive a perpetuity of €10 (including €4 compensation for
risk-taking). Their expectations are wrong, and they will actualb receive a
perpetuity of €15, providing a return of 15% instead of the 10%they anticipate.
In this case, we can reason that the opportunity cost of market investment is
either -
(a) a risk-adjusted rate equal to the required risk-free rate 6%, plus the
returns in excess of requirements (15% - lo%), i.e. 11% or
(b) a rate including compensation for risk-taking of the required return 10%
plus the excess return (1 5% - 10%) equals 15%. Note that the difference
between these rates is the risk premium of 4%. The opportunity cost is, in
both approaches, based on actual returns - which are, of course, unknown
at the time of taking the decision to invest. We proceed by using the rate
including compensation for risk-taking (i.e. ‘b’ above).
FIGURE 2
Y
Amount available
in Period 1
X
0 F Fl I C L Amount availablein
period 0
Even casual observation implies .that, during some periods, investors' expectations
change so dramatically that the general levels of market prices often cannot even
be approxiinations of the actual cash flows discounted at the market's required
rates of return. Consider the behaviour of prices on the New York and London
stock exchanges during the period January 1973 to June 1975. During the years
1973 and 1974 prices fell by almost 50% in New York and by about 70% in
London, then in November 1974 and the first four months of 1975 they rose by
50% in New York and 130% in London. Such major changes in a short period
could only derive from very significant revisions of formerly-held expectations.
The percentages are calculated in money terms but the changes are so marked
that inferences drawn from them would also seem valid in real terms.
Assume that
(1) throughout the period January to June 1975 the market expected a
return of 17% from investing in X Ltd.
(2) the company had l,OOO,OOO issued shares at 1st January 1975.
(4) the €750,000 raised under (3) was invested to yield 20% in perpetuity,
this being greater than the expected rate of 17%.
The stock market price per share in June could then be calculated as follows:
(a) (b)
Without transactions With transactions
3&4 3&4
Total Value EOOOs
Business existing 1st Jan
.6
(- x 1,000,000) 3,530(approx) 3,530(approx)
.17
New business
(”or) --
3,530
880
4,410
”
It is possible that the above illustrations are atypical because they are based on
an exceptional period. Merrett and Sykes’ (8) 1966 study implies otherwise.
Their results are summarised by the data in Appendix A and provide evidence
of continuing and large fluctuations in returns available from investment in
U.K. equities at different times. A different study by Barr (9) appears to
support rather than refute these conclusions so far as the U.K. is concerned.
Evidence concerning the annual monetary return to be obtained by investing
in the Standard and Poor’s Composite Index for all years from 1929 to 1972
is provided by Sharpe (10). He found a mean return of 10.64% with a standard
deviation of 21.06% and sizeable fluctuations in annual return occurred
throughout the period tested, so his data implies the existence of frequent
and marked revision of expectations by the U.S.A. market. It seems most
likely that very often the markets provide prices which are poor approxima-
tions of the actual future cash flows discounted at the rate of return
required by the market. The CAPM helps us to estimate the latter rate, and
conventional wisdom tells us that its use in capital budgeting will maximise
shareholders’ wealth.* This paper argues that such an outcome will only be
achieved if share prices,are adequate reflections of actual future cash flows
and expected rates of return, and that that is unlikely under the uncertainty
typically existing in practice. So it seems that the use of market expectations
derived figures of equity cost are unlikely to maximise shareholders’ terminal
wealth.
(a) It is presumably the rate which has been used in investor decisions
concerning investment in alternative markets, after taking account of required
compensation for the disutility of differential risk - and therefore probably
reflects opportunity costs in nonequity markets.
(b) One can argue that it should equal the marginal rate of time preference,
after compensation for risk-taking, of investors and therefore the firms’
shareholders (under the assumption that they are all currently investors rather
than borrowers). It should, therefore, help them to maximise the utility of
their consumption decisions over time if the rate is adopted.
(c) As it is the rate which they expect, provision of it should leave share-
holders reasonably satisfied and failure to provide it might create dissatis-
faction.
Following the above reasoning, we now define the objective of the firm as “the
maximisation of specified objectives, subject to providing shareholders with the
return which they would have required, at the date of expenditure of their funds,
from investments of equivalent risk”. This definition is compatible with a variety
of objectives, such as maximising the size of the firm, which can be viewed as
proxies for nonquantifiab1e.satisfactionsto be derived by participants in the
firm. A purpose of the project appraisal must now be to estimate the extent to
which considered projects meet the shareholders’ required return constraint. If
share prices were always good approximations of future cash flows and the rate
of return currently expected by shareholders from investments in the appropriate
risk class, the shareholder could sell his shares at any time to realise his expected
return. (Subject, of course, to the distorting effects of transaction costs.) In that
unlikely event, the market’s required return could be used with the satisficing
objective in capital budgeting without further assumptions. The real world
conditions of significant variability in share prices require, however, an additional
assumption to validate capital project appraisal techniques. Such an assumption,
now generally adopted in this paper, is that the firm’s existing shareholders will
hold their shares until the terminal date of the project, so that the firm could
distribute the total project net cash inflows to them. This relieves the analysis of
the problems deriving from ‘incorrectly’ priced shares. Within this framework,
the familiar discounted cash flow techniques can be used, subject only to their
well known limitations concerning reinvestment rates, with the selected criterion
rate.
Even if management adopts the satisficing objective, it has still tO estimate the
return required by investors from investments of the project’s risk class. Such
estimates require a model of market behaviour and, as discussed above, the
CAPM seems likely to be the most appropriate choice given the existing state of
knowledge. The latter, however, defines returns by reference to a single period.
(In practice “B’s” have usually been calculated for relatively short periods, e.g.
of one month’s duration.) Attempts to adapt the model to a multi-period setting”
often seem complex, requiring the input of data which is unlikely to be readily
estimated by management, yet not fully to utilise the information implicit in
current market data. Some improvement may be possible by simplifying the
procedure and recognising in the analysis the existence of a term-structure of
interest rates. The most casual observation reveals such a structure which, at any
point in time, yields different rates of interest for loans of different maturity. This
paper will not discuss the theories suggested to explain this phenomenon: it is
sufficient that it exists and has interesting implications for the use of the CAF’M,
e.g.
(a) a different ‘riskless’ rate may exist for each term (in this paper it is
assumed that it does) and
(b) for a long period, including a number of receipts of interest, the risk-less
rate can only be calculated under estimates of reinvestment rates during the
period; but such estimates are themselves presumably based on probability
distributions such that the rate is not ‘riskless’ after all!
Comment ‘b’ might be challenged on the basis that the interest rate is
contractually determined, so that it must be riskless. Such reasoning could
proceed by defining the rate as iIT in equation 3.
As all other terms are contractually agreed, iIT is also contractually agreed and
can be viewed as risk free. The CAPM is a single-period model, however, so the
where i3nt is the reinvestment rate in period t for the cash inflow in period n,
noting that the investor may face a different term structure of rates in each
period n.
i 2 is~the single period rate which solves the equation and other terms are as
previously defined.
T
I1 means multiplying the following expression (T-n) times, with the
t=n+1
variables taking on the assigned values.
Assuming that ijnt equals iIT, we can define the single-period “riskless’ rate of
return for a T-period term, h ~as ,
Insertion of this in equation 1 then gives a required return (over the T-period
term for a project of systematic risk class j) of
The reader will not that our analysis is still in the form of a single-period model.
Now defme R * 4 ~as j the periodic rate of return such that
Projects would meet the estimated required rate of return, defmed in equation 6,
if
We have, then, identified that we can validly use the multi-period rate R * 4 ~inj
Nw calculations under the assumptions
(a) that the reinvestment rate on periodic payments on ‘riskless’ T-term
securities is constant and equals the geometric mean of the single-period
return (after reinvestment) less one - i.e. that
-I
i 3 n t = ( I +izT)T- 1 for allyears
- so that the ‘riskless’ T-term rate is fully defined by contractual flows, and
The cost of equity, the CAPM and management objectives 113
(b) that the rate of return available on reinvestment in systematic risk class
j, of cash flows generated by projects of that risk class, is the same as the
selected discount rate - i.e. that rnt = R * 4 ~ for
j all periods.
These assumptions are unrealistic, but may provide as plausible a set of values as
can be produced by management forecasts. If that is the case they would appear
to be acceptable as a basis for practical decisions, when the NPV criterion rate
can be estimated using equations 6 and 7, i.e. it is
1
i4TF - 1
Obviously the estimates of many of the variables in the equation are likely to be
wrong - but this may still be the best approach available! The variables 4~ and
R m can ~ be estimated from data concerning present and past market transactions.
B ~ Tis more difficult to estimate because it is observed that the B applied by the
market to a firm’s shares changes over time. (W.F. Sharpe and G.M. Cooper have
provided an interesting paper on this topic, (16).) A natural starting point for
estimating B for a capital project may be the B’s calculated as applying to the
firm in the past. Management could then determine some rules to enable it to
adjust B, for project assessment, by reference to the perceived project risk class
(e.g. a replacement project in the firm’s major line of business might be assigned
the firm’s B). This procedure seems crude, but an even less satisfactory approach
has usually applied in the past, when no specific recognition has typically been
given t o systematic risk. Adoption of the assumptions specified earlier would
then allow the derived discount rate to be validly used in the furtherance of the
shareholder satisficing objective.
CONCLUSIONS
This paper claims that in the oncertain market environment the maximisation of
shareholders’ wealth is likely to be only achievable by luck and not by judgment,
because highly unrealistic assumptions, concerning both the market and the
individual shareholder’s alternative investment, are then necessary to the valid
use of the expected market rate of return with discounting models. The
maximising objective conventionally advanced in the literature of financial
management, therefore seems inappropriate as a proxy for the satisficing of
shareholders’ interests. An alternative, directly satisficing, aim is to provide
shareholders with the return which they require at the date of the investment,
i.e. the market’s expected rate of return. Such an approach seems fully consistent
with the concept of multiple corporate objectives and to require less unrealistic
assumptions than the wealth maximisation model when using capital project
evaluation methods. The analysis of the paper suggests a capital budgeting
application of the concepts of the CAPM whch recognises the term-structure of
interest rates when estimating the required rate of return for multi-period projects.
NOTES
lo E.g. See Bogue and Roll (3) and Bieman and Smidt (15) for discussions on the
application of the CAPM to multi-period capital project appraisal.
REFERENCES
N. Barr, “Real Rates of Return to Financial Assets since the War”, THE
THREE BANKS REVIEW, September 1975.
(14) E.F. Fama, “Efficient Capital Market: A review of Theory and Empirical
Work”, JOURNAL OF FINANCE, March 1973.
(15) H. Bierman and S. Smidt “Application of the Capital Asset Pricing model
to Multi-period Investments”, JOURNAL OF BUSINESS FINANCE AND
ACCOUNTING, Autumn 1975.
(16) W.F. Sharpe and G.M. Cooper, “Risk-Return Classes on New York Stock
Exchange Common Stocks, 1831-1967”, FINANCIAL ANALYSTS
JOURNAL, March-April 1972.
APPENDIX A
Returns from investment of lump sums in a U.K. stock exchange index in each
of years 1919-1956, each investment being held for ten years.
MC :Y TEI IS RE L TEN 1
YEARS 1919-29 130-45 946-56 9 19-2s 9 3 0 4 1946-56
Mean return 8.8% 5.5% 10.4% 11.4% 1.3% 6.2%
Standard devia-
tion of return 2.8% 2.8% 3.9% 4.0% 2.9% 4.4%
Range of return 2.9% 3 .O% 4.4% 3.6% - 1.9% -0.7%
to to to to to to
13.2% 0.5% 14.6% 17.2% 7.9% 11.3%
Greatest differ-
ence in return
from invest-
ments in
succeeding years 5.4% 4.4% 5.5% 6.3% 3.3% 5.5%
Proportionate
change in return
represented by
(d) above - i.e.
’d’/preceding
year’s rate 0.4 0.7 0.6 0.6 0.4 1.3
This appendix identifies assumptions under which the use of net present value
techniques to maximise shareholders’ wealth, would be valid if management
operated under certainty concerning the rate of return, to shareholders, from
investment alternative to that available in the firm. We proceed by assuming: -
2) That the term of the shareholders’ alternative equity investment will be the
same as that of the project under consideration by the firm, and that they will
reinvest intermediate cash flows prior to the terminal date.
3) That reinvestment rates available t o the shareholder and the firm are the
same.
5 ) That all cash flows occur at the end of the periods involved.
Under these assumptions the rate of return which the shareholder would obtain
on the investment alternative to a capital project would be “r” in equation 2
below
This equation defines the shareholder’s average annual return from the alternative
investment as the discount rate which equates the terminal value of the cash flow
associated with that investment with its present price. It will be noted that the
size of the rate “r” depends on the future rates of interest available, the size of
the cash flows (including present share price) and also the timing of those flows.
The average annual return actually obtainable from holding the alternative
investment is ‘rl ’ in the Equation A2.
Thus the shareholders’ opportunity cost is a function of the stream of cash flows
actually resulting from holding the alternative investment, and its present share
price.
The rate ‘r, ’ will only be constant for all shareholders if they all sell the identical
alternative investment at a common date, unless share prices are always an
adequate reflection of the future cash flows to be derived from holding the
shares and the future interest rates. They are unlikely to be such a reflection, so
a model which is based on an acceptance of share sale at different dates requires
assumptions which are as unreahstic as that of sale of shares at a common date,
and the latter seems tenable as the basis for analysis. This paper assumes that the
alternative investment would be sold at the terminal date of the capital
expenditure project, so that P, becomes A, and equations A1 and A2 are
equivalent. We continue on the basis of A l .
Po =-
(1 t r)"
E
t=1
At (1 t r ) n - t
a capital project will appear to be acceptable when evaluated at discount rate '?
if
i.e.
But Po equals I, so that on the stated assumptions At < Ct in each period if the
project has a positive or zero net present value at the discount rate ‘d’. In that
case, the terminal value of the project must be greater than, or equal to, that of
the shareholders’ alternative investment, regardless of the reinvestment rates
available, which are assumed to be identical for the shareholder and firm.
Similarly, a negative net present value will indicate that the project has a lower
net terminal value than the alternative, regardless of the reinvestment rates.
Although a correct accept/reject decision would emerge the NPV is not
necessarily a meaningful figure, however.