INT Campbell Direct.pdf

I
I
I
I
I
I
I 1
I 2
I 3
4
I 5
I 6
7
I 8
I 9
10
I 11
I 12
13
I 14
I 15
16
I 17
I 18
19
I
I
I
BEFORE THE IDAHO PUBLIC UTILITIES COMMISSION
INTERMOUNTAIN GAS COMPANY Case No. U-l034-88
PREPARED TESTIMONY OF RITCHIE A. CAMPBELL
Q. Please state your name and address for the record.
A. Ritchie A. Campbell, 21219 - 2nd Avenue S.E., Bothell,
Washington 98011.
Q. By whom are you employed?
A. I have been employed by the Uni versi ty of Washington as an
Assistant Professor of Finance since 1976.
Q. Please describe your employment and educational background.
A. I served in the United States Air Force from April of 1966 to
December of 1969. Upon discharge, I enrolled in the Univer-
sity of Washington's School of Business. I received my
Bachelor of Arts degree, cum laude, in Business Administra-
tion in 1972. My primary field of study was finance, and my
secondary fields were economics and accounting. I then
attended the University of pennsylvania's Wharton School to
continue my studies in finance and economics at the Masters
level. I received my Master of Business Administration
(M. B. A.) degree from the Wharton School in 1974. I majored
in finance while special izing in Portfol io and Capital Market
Theory. I was then admitted to the doctoral Program at
TESTIMONY - 1
I
I
I 1
2
I 3
I 4
5
I 6
I 7
8
I 9
I 10
11
I 12
I 13
14
I 15
I 16
17
I 18
I 19
20
I 21
I 22
23
I 24
I 25
26
I
I
I
Wharton. I received my Master of Arts (M.A.) degree in Busi-
ness and Applied Economics from the University of Pennsylvania's
Graduate School of Arts and Sciences in 1975, having completed
my doctoral areas in finance, economics, statistics, and
insurance. upon completion of my dissertation, The Demand
for Insurance: An Application of the Economics of Uncertainty
in 1977, I was awarded the Doctor of Philosophy (Ph.D.) in
Business and Applied Economics.
During my graduate studies (1972-1976), I studied under
Harry Markowitz, the pioneer of portfolio theory; Marshall
Blume and Irwin Friend, who have made many important contri-
butions to the capital asset pricing model and to the field
of finance; and Stephen Ross, a leader in the analysis of the
economics of uncertainty.
I was offered a faculty position at the University of
Washington in 1976. My teaching responsibilities have in-
cluded: Corporate Finance, Advanced Corporate Finance,
Advanced Portfol io Theory (CAPM), Personal Finance, and the
Economics of Uncertainty. I teach at the graduate (M.B.A.)
level and the doctoral (Ph.D) level.
Q. Have .you received any scholarships for high scholastic
achievement?
A. Yes. I was awarded a scholarship from Wharton to pursue my
M.B.A. studies. Then, upon admittance to the doctoral pro-
gram, I was designated as an S. S. Huebner Fellow. This
Fellowship contributed monthly stipends, book allowances, and
TESTIMONY - 2
I
I
I 1
2
I 3
I 4
5
I 6
I 7
8
I 9
I 10
11
I 12
I 13
14
I 15
I 16
17
I 18
I 19
20
I 21
I 22
23
I 24
I 25
26
I
I
I
tuition for the next three years to support my doctoral studies.
More recently, I was awarded University of Washington Faculty
research grants in 1977 and again in 1978.
Q. Are you currently involved in any research?
A. Yes. Most currently I am investigating "Corporate Financing
Decisions with Personal and Corporate Taxes". This research,
co-authored with Professor Charles Haley of the University of
Washington, is in a working paper format.
Additionally, I have been notified that a paper, devel-
oped from my dissertation, will be publ ished by The Journal
of Finance in a forthcoming issue, December of 1980.
Q. Are you a member of any professional societies?
A. I was initiated into Beta Gama Sigma, National Business
Honorary, for high scholastic achievement in 1972. I am also
a member of the American Finance Association and the Western
Finance Association.
Q. Have you delivered any professional speeches that have deal t
wi th the capital asset pricing model?
A. I have been invited to address the Seattle Society of Finan-
cial Analysis on two separate occasions; 1978 and 1979. Both
speeches dealt with various aspects of investment implications
of the capital asset pricing model, and portfolio theory.
Q. Dr. Campbell, have you previously testified as an expert witness?
A. Yes. I presented testimony to the Oregon Public utility Com-
mission (UF 3495) and to the Washington Publ ic Utility Commis-
sion (U-78-76) in 1979.
TESTIMONY - 3
I
I
I 1
2
I 3
I 4
5
I 6
I 7
8
I 9
I 10
11
I 12
I 13
14
I 15
I 16
17
I 18
I 19
20
I 21
I 22
23
I 24
I 25
26
I
I
I
Q. What was the nature of that testimony?
A. That testimony dealt with the determination of a fair rate of
return on common equity.
Q. Professor Campbell, what is your responsibility in this case?
A. I was asked to make a study and develop an opinion as to a
fair rate of return on common equity on behalf in Intermoun-
tain and to assemble into an exhibit those data on which I
have reI ied for such an opinion.
Q. What data did .you review in the course of your study?
A. I examined various books and records in Intermountain and
other published financial information.
Q. Is the exhibit which you prepared and relied on for your
opinion the exhibit that has been marked Exhibit 101 for
identification?
A. Yes.
Q. Was this Exhibit prepared by you?
A Yes.
Q. Is the information contained in Exhibit 101 true and correct
to the best of your knowledge and belief?
A. Yes.
Q. What standard is Intermountain using to determine its fair
rate of return on its total capital?
A. Intermountain is using the cost of capital standard to deter-
mine its overall fair rate of return. The cost of capital
standard indicates that there is no essential difference be-
tween the charges demanded by the firm's suppliers of capital
TESTIMONY - 4
I
I
I 1
2
I 3
I 4
5
I 6
I 7
8
I 9
I 10
11
I 12
I 13
14
I 15
I 16
17
I 18
I 19
20
I 21
I 22
23
I 24
I 25
26
I
I
I TESTIMONY - 5
I
I
I 1
2
I 3
I 4
5
I 6
I 7
8
I 9
I 10
11
I 12
I 13
14
I 15
I 16
17
I 18
I 19
20
I 21
I 22
23
I 24
I 25
26
I
I
I
dollar returns that just minimally satisfies its suppliers of
capi tal.
In capital budgeting theory, if a utility is allowed a
fair opportunity to earn. an overall fair rate of return, this
is equivalent to saying that the net present value of its
investments is expected to equal zero: that is, the present
value of the future allowed cash inflows is just expected to
equal the present value of the outflows (costs). Absent flo-
tation costs associated with the issuance of new financial
securities, if a firm has the opportunity to invest in posi-
tive net present value projects, the wealth of its common
stockholders will increase. But, this weal th increase would
be due to ratepayers making payments in excess of what was
minimally required. Conversely, if a firm was forced to
invest in projects whose net present value was expected to be
negative (the present value of the future cash inflows was
expected to be lower than the present value of the future
cash outf lows), stockholder weal th would decrease. But, this
decrease in stockholder weal th would be directly attributable
to ratepayers making payments for supplied services that were
less than what were minimally needed.
Thus, in the first case, ratepayers would be subsidizing
stockholders; and in the second case, the stockholders would
be subsidizing ratepayers (stockholder wealth is being con-
fiscated). Neither of these cases are equitable. These
potential inequities would not occur if the firm was expected
TESTIMONY - 6
I
I
I 1
2
I 3
I 4
5
I 6
I 7
8
I 9
I 10
11
I 12
I 13
14
I 15
I 16
17
I 18
I 19
20
I 21
I 22
23
I 24
I 25
26
I
I
I
to invest in projects whose net present value was zero: rate-
payers would be paying the minimally fair dollar rates and
the firm's suppliers of capital could expect to receive their
minimally acceptable rates of return. This is the goal of
utility regulation. To achieve this goal all future costs
are estimated--including the cost of supplied capital--and
rates are designed such that the firm can be expected to gen-
erate future cash inflows that just cover these costs. In
terms of common stock, and abstracting from flotation costs,
what one must estimate is the firm's common stockholders'
minimally acceptable rate of return.
It should be clear that what has been stated was on an
expectation basis. Many conditions, all presently uncertain,
will affect the firm's cash flow generation in the future as
well as its suppl iers of capital's minimal demands. Therefore,
a firm's overall rate of return, previously determined to be
fair, must be periodically updated as the future unfolds and
past allowed rates become either too high or too low.
Q. Does a firm's stockholders' minimally acceptable rate of
return measure the firm's fair rate of return on common equity?
A. No, not precisely. Due to flotation costs, a utility's
stockholder's minimally acceptable rate of return is less
than the firm's fair rate of return on common equity.
Direct flotation costs incurred by any firm raising new
funds in the capital market include investment banking fees,
attorney and accounting fees, printing fees and the like.
TESTIMONY - 7
I
I
I 1
2
I 3
I 4
5
I 6
I 7
8
I 9
I 10
11
I 12
I 13
14
I 15
I 16
17
I 18
I 19
20
I 21
I 22
23
I 24
I 25
I
26
I
I
More indirect costs include discounts from face value in the
case of debt and preferred stock or discounts from current
market prices in the case of common stock (underpricing).
Additionally, in the case of common equity, another indirect
cost is called "pressure", which is really the dilution
effect on market prices due to the issuance of additional
common stock.
Q. Are flotation costs associated with the issuance of debt and
preferred stock allowed to be recovered in utility rate hearings?
A. Yes. They are recovered directly through the calculation of
their respective embedded costs.
Q. Are the flotation costs associated with the issuance of com-
mon stock allowed to be recovered in utility rate hearings?
A. Most generally, yes. It would be inconsistent to rightfully
allow a utility to recover its flotation costs associated
with debt and preferred stock, yet not allow the costs asso-
ciated with common stock.
Q. Dr. Campbell, will you explain how a utility's fair rate of
return on its long-term debt is calculated and how this
calculation takes into account the flotation costs associated
with the issue?
A. A utility's fair rate of return on its long-term debt is nor-
mally called its embedded cost of debt. The embedded cost of
debt is defined as that rate of interest that makes the pres-
ent value of the future dollar obI igations associated with
the issuance of long-term debt (future semi-annual interest
TESTIMONY - 8
I
I
I 1
2
I 3
I 4
5
I 6
I 7
8
I 9
I 10
11
I 12
I 13
14
I 15
I 16
17
I 18
I 19
20
I 21
I 22
23
I 24
I 25
26
I
I
I
payments and face value upon maturity) equal in present value
to the funds that the firm received from the debt issue (face
value issued less flotation costs and discounts). It is use-
ful to note that this rate forces the net present value of
the utility's long-term debt to equal zero. When applied to
the debt component of rate base the embedded cost of debt
produces the minimum dollars needed to support the firm's
financial obligations associated with the debt issue.
Another way to illustrate this is by noting that if the
utility deposited in a bank the proceeds it received from the
debt issue, and if the bank paid on an annual compounded basis
the embedded cost, then the util i ty could make periodic wi th-
drawals from its bank account to meet its legally required
payments (periodic interest) and upon maturity, having made
the last legally required interest payment as well as prin-
c ipal repayment, the balance in the bank account would equal
zero. Thus, a compounded rate of return equal to the firm's
embedded cost of debt, earned on the debt component of rate
base, will just provide the necessary future dollars that
minimally satisf ies its legally required payments.
Q. Will you now explain how a utility's fair rate of return on
preferred stock is determined and how this calculation takes
into account the flotation costs associated with the issue?
A. The determination of a fair rate of return on preferred stock
is analogous in concept to the determination of fair rate of
return on the utility's long-term debt. That rate of interest
TESTIMONY - 9
is found that equals the present value of the net proceeds
the firm received from the preferred stock issue (face value
less discounts and expenses) with the present value of its
financial obI igations (quarterly preferred dividend payments).
The rate of interest that creates a zero present value
is again deemed to be a fair rate of return because its appli-
cation to the preferred stock component of rate base results
in the minimum dollars required to meet the firm's future
financial obligations associated with the issuance of pre-
ferred stock.
The embedded costs of preferred and long-term debt both
increase with the coupon rate, the selling discounts and with
the amount of flotatio~ costs. Absent any discounts and if
flotation costs equalled zero, the fair rate of return on
long-term debt and preferred stock would equal their coupon
rates.
Q. You have discussed the financial standards for a fair rate of
return on a utility's long-term debt and its preferred stock.
What is the .financial standard for a fair rate of return on
common equity?
A. A utility can be said to earn a fair rate of return on the
common equity component of its rate base if the dollars that
are expected to be produced enable the utility's current and
potential stockholders to just expect to earn their minimally
acceptable rate of return on the market value of their invest-
ment. Due however to flotation costs associated with common
TESTIMONY - 10
I
I
I 1
I
2
3
I 4
I
5
6
I 7
I 8
9
I 10
I
11
12
I 13
I 14
15
I 16
I 17
18
I 19
I 20
21
I 22
I 23
24
I 25
I 26
I
I
stock financing, a utility's current stockholders can only
expect to earn their minimally acceptable rate of return if
the util i ty is allowed to earn a return that is slightly
greater than the stockholders' minimal demands.
Q. If the firm's fair rate of return is greater than its stock-
holders' minimal demands, does this mean that the stock-
holders can expect to earn more than their minimally demanded
rate of return?
A. Absolutely not, and the reason is exactly analogous to the
case of long-term debt and preferred stock. If there were no
flotation costs, the firm's fair rate of return on common
equity would exactly equal its stockholders' minimally
demanded rate of return. Just as, in this case, the firm's
fair rate of return on its long-term debt would just equal
its coupon rate. But, due to flotation costs, the firm must
earn on its common equity component of rate base an amount
greater than its stockholders' minimal demands before the
stockholders can just expect to earn their minimal demands.
Q. What is the resulting outcome of the price of a util i ty' s
common stock when the regulatory procedure of applying the
allowed rate of return to the equity component of rate base
is used?
A. I have shown in Schedule 1, using two different methods, that
stock market prices for a utility's common stock will have a
tendency to equal the book val ue of common stock if: (1) the
allowed rate of return is a precise estimate of the firm Is
TESTIMONY - 11
I
I
I 1
2
I 3
I 4
5
I 6
I 7
8
I 9
I 10
11
I 12
I 13
14
I 15
I 16
17
I 18
I 19
20
I 21
I 22
23
I 24
I 25
26
I
I
I
stockholders' minimally acceptable rate of return, and if (2)
flotation costs do not exist, and if (3) the firm has a fair
opportuni ty to earn this allowed rate.
However, if any of the preceding three conditions are
violated, market val ues can be expected to be less than book
values. For example, when an analyst estimates a utility1s
stockholders' minimally acceptable rate of return, or when-
ever a judgment is called for and the most conservative esti-
mate is made, the stockholders' minimal dollar demands are
not likely to be attainable. Moreover, even if the stock-
holders' minimally acceptable rate of return were accurately
estimated, but a flotation cost adjustment was not made, the
stockholders' minimal dollar demands would again not be
attainable. Finally, even if the stockholders' minimal
demands were accurately estimated, and even if an accurate
flotation cost adjustment was made, but the estimates of the
utility's operating costs were understated and/or the reve-
nues were overstated, then again, the stockholders could not
expect to attain their minimally acceptable rate of return at
a market price equal to the firm's book value. That is, if
the firm's fair rate of return was accurately estimated, but
the firm was not allowed a fair opportunity to actually earn
this fair rate of return, the utility's current stockholders
could not expect to earn their minimal demands.
Q. Dr. Campbell, what are the consequences of a utility's
stockholders not expecting to earn their minimally acceptable
TESTIMONY - 12
I
I
I
1
2
I 3
I
4
5
I 6
I
7
8
I 9
I
10
11
I 12
I
13
14
I 15
I
16
17
I 18
I
19
20
I 21
I 22
23
I 24
I 25
26
I
I
I
rate of return?
A. If any of the three previously listed conditions are violated,
ratepayers have paid less than fair prices and stòckholders
cannot expect to earn their minimally acceptable rate of
return; the market price of the util i ty' s common stock will
fall below its book value and will continue to fall until a
new equil ibrium price is reached. At this new equil ibrium
price, new stockholders can expect to earn their minimally
acceptable rate of return, but the original stockholders will
face a loss of wealth. Associated with this loss (confisca-
tion) of stockholder wealth is the undesirable implication
that, if the firm issued new common stock, additional stock-
holder loss would continue to occur. Management, recognizing
this principle, would be reluctant to issue new common stock
in order to meet its financial requirements. The firm would
then be faced with two choices: it could meet its financial
requirements by issuing additional debt, or it could curtail
its planned expenditures.
Regarding the first choice, if additional debt were
issued, the firm would become riskier: a greater proportion
of its earnings would be subject to fixed charges so that the
range of safety over which future earnings could temporarily
fall without jeopardizing the solvency of the firm would
decrease. Moreover, because of the increased leverage asso-
ciated with debt, small changes in the firm's earnings would
lead to large changes in the volatility of the residual earn-
TESTIMONY - 13
I
I
I
1
2
I 3
I
4
5
I 6
I
7
8
I 9
I
10
11
I 12
I
13
14
I 15
I 16
17
I 18
I 19
20
I 21
I 22
23
I 24
I 25
26
I
I
I
i ng s that the common stockholders expect to ~ece i ve . Thus,
from the common stockholders' perspective, the issuance of
addi tional debt makes the common stock riskier and the stock-
holders' minimally acceptable rate of return would increase.
Associated with the increased minimal demands, common stock
prices could be expected to again fall and the firm would be
in a less favorable position to issue new common stock in the
future. Thus, the firm would be forced into a more costly
capi tal structure that ratepayers would ultimately bear.
Regarding the second choice that a firm has if its
stockholders cannot expect to earn their minimally acceptable
rate of return, if the firm decreases its financial needs by
postponing needed plant expenditures, then more efficient
technology, which could have been introduced, would not be
forthcoming. The result would be greater future ratepayer
costs than those that would have been necessary if the past
needed investments were made.
The electric industry's inabil i ty to fulf ill future
energy demands has been well documented in the states of
Idaho, Washington and Oregon. Outright moratoriums on new
customer hook-ups, or hook-ups at high costs have been and
are being explored. The result of these current moratoriums,
as well as the difficulty of building nuclear plants for
future generation of energy, dictates an increasing role for
alternative sources of energy. The burden of meeting future
energy demands in the Northwest will fall most heavily on the
TESTIMONY - 14
I
I
I 1
2
I 3
I 4
5
I 6
I 7
8
I 9
I 10
11
I 12
I 13
14
I 15
I 16
17
I 18
I 19
20
I 21
I 22
23
I 24
I 25
26
I
I
I
natural gas industry. Intermountain recognizes its impor-
tance in providing energy to the residents of Idaho. Inter-
mountain, however, also recognizes that increased. plant
investment is needed now to meet its responsibility in the
future. Commensurate with this increased plant investment
are near-term capital market needs. It is because of these
needs that the determination of a fair rate of return is so
serious a matter. Any ratepayer "savings" associated with a
utility earning less than its fair rate of return are illusory.
The inevitable fact would be loss of current stockholders
weal th, a curtailment of current investment needed to supply
energy in the future or a costly capital structure that rate-
payers must ultimately bear.
All of us at these proceedings are ratepayers. All of
us recognize the fact that "cheap" (incorrectly priced) energy
does not solve problems; it creates problems for the future.
By correctly pricing energy, a greater proportion of our
budgets must be used for energy; however, if energy is not
correctly priced today, future energy may be sporatic or
perhaps, nonexistent.
Q. You have discussed the concepts of a fair rate of return on a
utility's debt, preferred stock and common stock and the
necessi ty of accurately estimating all factors that affect
whether stockholders can reasonably expect to earn their
minimally acceptable rate of return. What methods did you
use to estimate Intermountain's fair rate of return on its
TESTIMONY - 15
I
I
1
I 2
I 3
4
I 5
I 6
7
I 8
I \ 9
10
I 11
I 12
13
I 14
I 15
16
I 17
I 18
19
I 20
I 21
22
I 23
I 24
25
I 26
I
I
I
common equity?
A. I used three different methods to determine Intermountain's
fair rate of return on its common equity. The first method
is a risk premium method which recognizes that common stock
investing is riskier than investing in corporate AAA bonds,
and as such, common stockholders require a greater rate of
return. The second method is derived from current capital
market theory: it is called the CAPM (capital asset pricing
model) approach. And the third method is called the DCF
(discounted cash flow) approach.
Using these methods, I estimated Intermountain' s stock~
holders' minimally acceptable rate of return. In order to
determine its fair rate of return on common equity, I multi-
plied the stockholders' minimally acceptable rate of return
times the flotation cost adjustment factor of 1.06. As I
have mathematically shown in Schedule 2, equation (5), this
adj ustment is necessary to translate the stockholders' require-
ments into a fair allowed rate of return to be applied to
intermountain's equity component of rate base.
Q. How do you use equation (5) of Schedule 2 to arrive at a 1.06
flotation cost adjustment for Intermountain?
A. Most analysts assume that flotation costs of at least 10%
conservatively represents the percent of total issue costs
incurred when common equity is issued. Hence, the factor
1.11 in equation (5):1(1-.1 .
Intermountain's dividend payout ratio has been, on aver-
TESTIMONY - 16
I
I
1
I 2
I 3
4
I 5
I 6
7
I 8
I 9
10
I 11
I 12
13
I 14
I 15
16
I 17
I 18
19
I 20
I 21
22
I 23
I 24
25
I 26
I
I
I
age, about two-thirds of its earnings, and this is very rep-
resentative of other utili tys' payout ratios. Substituting
. 67 for the value "d" in equation (5) solves for fhefact
that Intermountain's fair rate of return on its common equity
is 1.06 times its stockholders' minimally acceptable rate of
return.
Q. Dr. Campbell, why did you use so many different methods to
estimate Intermountain's stockholders 'minimally acceptable
rate of return?
A. I used three different methods that use different data input
in order to check the internal consistency of my findings.
This is particularly important in the area of estimating a
firm's cost of common equity because the analyst's estimate
translates into rather substantial revenue requirements for
the firm. Therefore, one must be very sensitive as to the
reasonableness of the value that is being projected. If only
one approach is used, there is no frame-of-reference regarding
the reasonableness of the estimate except a historical frame-
of-reference of what was found to be reasonable in the past.
But, using a historical frame-of-reference in our current
financial markets would not provide accurate insight into
what is currently reasonable.
For example, all of us recognize that a supplier of labor
to a firm must be compensated for the rate of inflation in
his wages just to "break-even." The same is true for a
supplier of capital; the minimum compensation must, at a
TESTIMONY - 17
minimum, change with the inflationary changes--otherwise the
supplier of capital would find that he would be better off
consuming his capital today rather than foregoing current
consumption opportunities by supplying his capital to the
firm. using just this one component of what determines an
investor's minimally acceptable rate of return--inflation--it
is useful to evaluate how inflation has changed most recently
in the United States. The first half-decade of the 1960' s
(1960-1964) was characterized by a 1.2% rate of inflation;
the second half-decade of the 1960' s was characterized by a
3.8% rate of inflation. The first half-decade of the 1970' s
was characterized by a 6.7% rate of inflation; the second
half-decade of the 1979's was characterized by an 8.1% rate
of inflation. Even more dramatic is the fact that inflation
was 13.3% last year, and for the first three months of this
year inflation has been in excess of 18%.
This means that an equivalent, inflation-adj usted, fair
rate of return must currently be 7 to 10 percentage points
higher than those average rates found to be fair during the
first half-decade of the 1970' s (with an average inflation of
6.7%). Of that, an equivalent, inflation-adjusted, fair rate
of return must currently be 5 to 8 percentage points higher
than those average rates found to be fair during the second
half-decade of the 1970' s.
It is because of this dramatic change of inflation in
our economy that an unadjusted, historical frame-of-reference
TESTIMONY - 18
I
I
I
1
2
I 3
I
4
5
I 6
I
7
8
I 9
I
10
11
I 12
I 13
14
I 15
I 16
17
I 18
I 19
20
I 21
I 22
23
I 24
I 25
26
I
I
I
is rendered obsolete for rate setting in regulatory matters,
for price setting in competitive markets, for wage negotia-
tions in labor markets, and for required yields in our current
money markets. The fact that a historical frame-of-reference
may not provide meaningful insight into the reasonableness of
a return requirement was noted by the Supreme Court in 1923.
The Court stated that, "A return may be reasonable at one
time and become too high or low by changes affecting oppor-
tunities for investment, the money market, and business con-
ditions generally." Bluefield Waterworks & Imp. Co. vs.
Public Service Commission of West Virginia, 262 u.s. 679, 693
(1923). (Emphasis added.)
Q. Dr. Campbell, before y~u describe the approaches that you
used to estimate Intermountain's stockholders' minimally
acceptable rate of return, will you briefly summarize how a
minimally acceptable rate of return is determined by an
investor as opposed to how it is estimated by an analyst?
A. Yes. Perhaps the eas ies t explanation of how an inves tor
determines his minimally acceptable rate of return starts by
noting that this rate has two important component parts:
( 1) A current opportunity rate of return on a financial
securi ty whose future cash flows are certain, and
(2) An additional component to compensate the investor
if the financial security has uncertain future cash
flows.
Regarding the first component: a financial security
TESTIMONY - 19
I
I
1
I 2
I 3
4
I 5
I 6
7
I 8
I 9
I
10
11
I 12
I
13
14
I 15
I
16
17
I 18
I 19
20
I 21
I 22
23
I 24
I 25
26
I
I
I
whose future cash flows are certain is generally called a
r iskfree security. And, the rate of return associated with
investing in a financial security whose future cash flows are
certain is called the riskfree rate of return.
Regarding the second component: the additional compen-
sation (in excess of the riskfree rate) that investors require
before they can be expected to divert capital away from risk-
free investing into, say, common stock investing is generally
called the risk premium in economics.
Thus, we can conclude that an investor's current mini-
mally acceptable rate of return is determined by the currently
available riskfree rate of return, plus a minimum risk pre-
mium that is necessary to entice investors to invest their
capi tal in firms that generate uncertain future returns.
Q. Based on what you have said, Dr. Campbell, would it be fair
to say that two different firms with different degrees of
uncertainty in their ability to generate future returns would
have different minimally acceptable rates of return?
A. Yes, precisely. The current riskfree rate of return is the
same for all firms. The investors' perception of the uncer-
tainty of the firm's ability to generate future cash flows,
however, may differ for different firms. Hence, the required
minimally acceptable rate of return will differ due to dif-
ferent demanded risk premiums.
Q. Is there any other way to view how an investor determines
what minimum compensation he requires on a firm's common stock?
TESTIMONY - 20
I
I
I 1
2
I 3
I
4
5
I 6
I
7
8
I 9
I 10
11
I 12
I 13
14
I 15
I 16
17
I 18
I 19
20
I 21
I 22
23
I 24
I 25
26
I
I
I
A. Yes. The component parts of how an investor determines his
minimally acceptable rate of return may be expanded and made
more specific. The factors that affect an investor's mini-
mally acceptable rate of return are:
(1) the investor's impatience to consume,
(2) the investor's expectation of the future rate of
inflation,
(3) the investor's uncertainty regarding the future
rate of inflation, and
(4) the investor's uncertainty regarding the future
cash flows that the common stock will produce
(future dividends and market prices).
Because the fourth factor does not come into play in the
determination of the investor's minimally acceptable rate of
return for riskfree investing (the future cash flows are pre-
cisely known), we can conclude that the investor's minimally
acceptable rate of return corresponding to riskfree investing
is determined by the first three factors. Thus, the greater
the investor's impatience to consume, the greater his expec-
tation of future inflation, or the greater his uncertainty
regarding the future inflation rate, then the greater is the
minimally acceptable rate of return on riskfree investing.
Given the unprecedented rate of inflation in our economy
and the current uncertainty of about its future rates, it is
clear that current minimally acceptable rates of return on
riskfree investing are very high. Moreover, given the fact
TESTIMONY - 21
I
I
I 1
2
I 3
I 4
5
I 6
I 7
8
I 9
I 10
11
I 12
I 13
14
I 15
I 16
17
I 18
I 19
20
I 21
I 22
23
I 24
I 25
26
I
I
I
that, in addition to the riskfree rate, investors require an
additional premium before they will invest their capital in
common stock that produces uncertain future returns, the
current minimally acceptable rates of return on common stock
investing is also very high.
Q. Professor Campbell, will you now describe the process that an
analyst takes to estimate an investor's minimally acceptable
rate of return?
A. Yes. An analyst can either estimate the investor's minimally
acceptable rate of return directly or indirectly.
A direct estimate is obtained by: (1) estimating the
riskfree rate and (2) adding to this estimate the estimate of
the risk premium, commensurate with the risk of the firm's
common stock, that investors seem to require before they will
acquire the firm's common stock.
An investor's minimally acceptable rate of return can
also be estimated indirectly by: (1) evaluating current prices
for the financial security, (2) estimating the investor's
projection of future cash flows associated with the financial
securi ty and (3) mechanically (indirectly) solving for the
inves tor's impl ied minimally acceptable rate of return.
The DCF approach is an indirect approach; it does not
attempt to directly measure the component parts of an inves-
tor's minimally acceptable rate of return. Therefore, it
does not attempt to quantify the relative risk of the security
in order to determine the appropriate risk premium.
TESTIMONY - 22
I
I
I 1
2
I 3
I 4
5
I 6
I 7
8
I 9
I 10
11
I 12
I 13
14
I 15
I 16
17
I 18
I 19
20
I 21
I 22
23
I 24
I 25
26
I
I
I
CAPM is a direct approach; it directly measures the com-
ponent parts of the investor's minimally acceptable rate of
return: the riskfree rate, the relative level of .securi ty
risk, and the minimum compensation per unit risk that inves-
tor's demand. Indeed, the CAPM approach may be thought of as
an improvement over the simpliest and perhaps older approach
used in regulatory hearings to estimate a utility stockholder's
minimally acceptable rate of return--the risk premium approach.
For example, the risk premium approach starts by noting
that investors, rather than investing in common stock, could
have invested in corporate AAA bonds. Because the investment
in corporate AAA bonds is safer than common stock investing,
one must add to the current AA yield a risk premium to
determine the stockholders' current minimally acceptable rate
of return.
Q. Dr. Campbell, have you undertaken a study to determine the
risk premium that investors seem to require on common stock
investing relative to AAA corporate bond investing?
A. Yes. Using data on corporate bond AAA yields extracted from
Moody's Bond Record from 1950 through 1979 (these data are
contained in Schedule 3 of Exhibit 101), and data from common
stock investing (contained in Schedule 4 of Exhibit 101), I
found that, on average, common stockholders require about 7%
( 6.9%) more than what they could have achieved through
investments in AAA corporate bonds.
Q. What yields can investors expect to receive on AAA. corporate
TESTIMONY - 23
I
I
I 1
2
I 3
I 4
5
I 6
I 7
8
I 9
I 10
11
I 12
I 13
14
I 15
I 16
17
I 18
I 19
20
I 21
I 22
23
I 24
I 25
26
I
I
I
bonds over the next three years?
A. Data Resources, Inc. 's well respected "U.S. Forecast Summary"
for April, 1980, is predicting average rates of 13.08%,
11.61% and 11.96% over the years 1980, 1981 and 1982. The
average of these rates over the next three years, 12.2%, is a
reasonable value that investors should expect.
Adding an average risk premium requirement of 6.9% to
the average AAA rate of 12.2% results in an implied near term
stockholder requirement of 19.1%.
Q. Do you find this rate of return excessive?
A. Absolutely not. I must admit, though, that it has taken me
some time to adj ust my thinking to the current rates in our
securi ties markets. And, I expect that for most of us the
reality of our current environment has not been fully assimi-
lated into our day-to-day thinking. But let me describe some
current financial conditions.
A. First of all, it is becoming progressively clearer
that inflation is likely to equal 16% for the year
1980. The reasons I have formed this opinion follow:
1. 1979' s rate of inflation was 13.3%.
2. Eval uating the most recent months of January,
February and March of 1980, the CPI index
increased at a monthly rate of 1.4% during
each month. This translates to an 18.2%
annual rate of inflation. So for one-fourth
of this year, the rate of inflation has been
TESTIMONY - 24
I
I
I 1
2
I 3
I 4
5
I 6
I 7
8
I 9
I 10
11
I 12
I 13
14
I 15
I 16
17
I 18
I 19
20
I 21
I 22
23
I 24
I 25
26
I
I
I TESTIMONY - 25
at an 18.2% rate.
3. The producer price index figures, released for
the months of January and February show
annualized increases in excess of 24%. Because
these cost increases have not been reflected
in consumer prices yet, their impact on the
CPI will not occur for three or four months.
4. President Carter has mandated a 10~ per gallon
increase in the price of gasoline. Not only
will this directly affect the CPI because of
direct gasol ine sales to ind ividual consumers,
it will indirectly affect the CPI index because
it represents another increase in costs that
businesses face . These add itional costs will
be reflected in the price of their products.
5. For the first two months, January and February
of 1980, food has not increased in price, and
the component of food prices in the CPI has
remained unchanged. But farmers will have to
increase their prices soon because their costs
are accelerating. Food prices will increase
this year.
6. The President and the Federal Reserve System's
new anti-inflation measures are inherently
inflationary. Due to the Federal Reserve
System's increase in the discount rate over the
I
I
1
I 2
I 3
I
4
5
I 6
I
7
8
I 9
I
10
11
I 12
I
13
14
I 15
I
16
17
I 18
I 19
20
I 21
I 22
23
I 24
I 25
26
I
I
I
TESTIMONY - 26
B.
past six months, interest rates are extremely
high. Moreover, they are expected to continue
to stay at relatively high rates for sometime.
Because of these high rates, businesses must
again increase their product prices. And
these have not been fully reflected in the CPI
as. of yet.
7. It should be clear that President Carter's well-
publicized election year balanced budget pro-
mise will not affect this year's anticipated
deficit of 50 to 75 billion dollars. Indeed,
the budget mayor may not balance in fiscal
1981. And, even if it does, there is already
talk of using the "excess" funds for personal
tax relief. This, of course, would itself be
inflationary as consumers start a new spending
spree.
For all of these reasons, it would not be a
surprise to me if we actually have a rate of infla-
tion for 1980 that is 16% or greater. And, I would
be amazed if it were as low as 14 %; this just does
not seem to be obtainable. My most conservative
estimate is a 14 1/2% overall rate of inflation
this year, but I am starting to suspect that this
number might be too low.
Current money market yields bear out this
I
I
1
I 2
I 3
I
4
5
I 6
I
7
8
I 9
I
10
11
I 12
I
13
14
I 15
I
16
17
I 18
I 19
20
I 21
I 22
23
I 24
I 25
26
I
I
I
estimate of inflation for 1980. As reported in The
Wall Street Journal (April 18, 1980 and May 5, 1980):
1. The prime rate, the rate charged tó the most
credi tworthy businesses, has ranged from
18 1/2% to 20 % .
2. The marg in rate on common stock investing has
been 18% to 20%.
3. 90-day commercial paper unsecured notes sold
by the most creditworthy businesses, have sold
at 12.25% to 16% rates.
4. One-year certif icate of depos its have paid
11 1/8% to 14 % .
5. Six-month money market rates, paid by banks,
have yielded 10.8% to 13.6%, and money market
funds have yielded 23.4% to 18.2%.
6. One-year maturing u. S. Treasury notes have
been selling to yield 10.25% to about 15%.
7. Six-month Treasury bills have traded in a
range from 11.7% to 15%.
Given the seriousness of the rates of inflation in our economy,
and given the rates of return investors can currently earn on
very safe alternative investment media, a 19% rate require-
ment on common stock investment, rather than appearing exces-
sive, may not provide sufficient incentive for investors to
transfer their funds into common stock investing.
Q. Dr. Campbell, you mentioned that you also estimated Inter-
TESTIMONY - 27
I
I 1
I 2
I 3
4
I 5
I 6
7
I 8
I 9
10
I 11
I 12
13
I 14
I 15
16
I 17
I 18
19
I 20
I 21
22
I 23
I 24
25
I 26
I
I
I
mountain's stockholders' minimally acceptable rate of return
using the capital asset pricing model, is that" correct?
A. Yes.
Q. Does the capital asset pricing model provide a direct esti-
mate of the stockholders' minimally acceptable rate of
return, or is it an indirect approach like the discounted
cash flow method?
A. The capital asset pricing model (CAPM) is a risk premium
approach: It provides a direct estimate of the four component
parts of an investor's minimally acceptable rate of return.
The CAPM formula for estimating a firm's stockholders' mini-
mally acceptable rate of return (equation (2) of Schedule 8
of Exhibit 101) states that investors require, in addition to
the riskfree rate of return, a risk premium that may vary for
different firms depending upon the volatility of their common
stock returns.
Because the capital asset pricing model specif ies the
measurement of risk of the firm's common stock from the point
of view of investors, and specif ies the reward investors
should expect to receive for bearing this risk, CAPM repre-
sents a vast improvement over the older risk premium measures.
Indeed, the capital asset pricing model probably represents
the single most important advance in the field of finance
over the last three decades. No university that I am aware of
teaches finance without presenting the capital asset pricing
model: it has permeated both corporate finance and personal
TESTIMONY - 28
I
I
1
I 2
I 3
4
I 5
I 6
7
I 8
I 9
10
I 11
I 12
13
I 14
I 15
16
I 17
I 18
19
I 20
I 21
22
I 23
I 24
I
25
26
I
I
I
finance textbooks.
Q. Dr. Campbell, before you present your estimate of Intermoun-
tain's cost of common equity using the CAPM approach, will
you summarize, briefly, the historical development of the
capi tal asset pricing model?
A. Current capital market theory rests upon two major corner-
stones: efficient capital markets and portfolio theory. To
describe the historical development of the capital asset
pricing model, therefore, one must first describe the notion
of efficient capital markets and the concept of portfolio
theory.
Q. What is an efficient capital market?
A. From an economic perspective, the primary function of a capi-
tal market is to eff iciently channel scarce capital from
saving units (investors) to consuming units (firms). This
intermediation function is said to be efficient if scarce
capital is allocated to the most productive users of capital.
Market rates of return on a firm's stock provide signals
to investors that identify potentially eff icient users of
capi tal. In an eff icient capital market, the firm whose
security provides the highest market return, for a given
level of risk, is the most eff icient user of capital. That
is, in efficient capital markets, the return on a firm's
stock provides a correct signal to investors concerning its
real productive opportunities. A market rate of return sig-
nal will be a correct estimate of the firm's real productive
TESTIMONY - 29
I
I
1
I 2
I 3
4
I 5
I 6
7
I 8
I 9
10
I 11
I 12
13
I 14
I 15
16
I 17
I 18
19
I 20
I 21
22
I 23
I 24
25
I 26
I
I
I
opportuni ties if information pertaining to the firm's produc-
tive opportunities is quickly and accurately incorporated
into its stock market price. But this means that. firms with
the same amount of investor perceived risk, but with dif-
ferent productive opportunities, will differ in market prices
solely because of the productive differences. If stock prices
already fully reflect different firms' productive abilities
(they are not underpriced nor overpriced on average), then
investors can expect to receive neither more nor less than a
future market rate of return that provides minimal compen-
sation for the "time value of money" (the riskfree rate) and
for the risk inherent in the investment.
Contrastingly, a market is ineff icient if the stock
market return signal on a firm's stock is an incorrect de-
scription of the firm's real productive opportunities. The
market is inefficient in this case because investors, acting
on the incorrect market signal and attempting to maximize
the ir wealth, will shift capital out of potentially produc-
tive firms and into a firm that has a high stock market return
but, perhaps, no real productive opportunities. In this
case, scarce capital is not allocated to the most productive
user of capital.
The overriding impression that one is left with, after
reading numerous capital market studies, is that our capital
markets are highly competitive and efficient marketplaces.
Q. From an investor's perspective, what implication arises from
TESTIMONY - 30
I
I
I 1
2
I 3
I 4
5
I 6
I 7
8
I 9
I 10
11
I 12
I 13
14
I,15
I 16
17
I 18
I 19
20
I 21
I 22
23
I 24
I 25
26
I
I
I
the fact that our capital markets are efficient?
A. Given that investors, on the average, can just expect to be
compensated for the time value of money and the risk inherent
in the security in an eff icient capital market, most investors
would be better off if they: (1) selected a desired level of
risk and (2) formed a portfolio of securities that is con-
sistent with this desired risk level.
Q. Why should investors hold portfolios?
A. Past empirical studies have demonstrated that individual
securi ties are correctly priced, on the average, in our stock
markets. Therefore, in order to increase the possibility of
actually receiving a correct rate of return, investors should
allocate their investable funds among many different securi-
ties. In so doing, if a few securities in the portfolio
actually provide returns lower than what was correct, other
securities in the portfolio may actually provide returns that
were higher than correct returns. If investors hold port-
folios, these individual discrepancies will have a tendency
to offset one another. Therefore, if investors hold port-
folios consisting of many securities, they are increasing the
probability that their actual portfolio returns will be
correct returns.
The fact that most investors should hold well diversified
portfolios was first pointed out by Harry Markowitz, the pio-
neer of modern capital market theory. Markowitz noted, in
the 1950's, that, except in the unlikely case when all secur-
TESTIMONY - 31
I
I
I
1
2
I 3
I
4
5
I 6
I 7
8
I 9
I 10
11
I 12
I
13
14
I 15
I
16
17
I 18
I
19
20
I 21
I
22
23
I 24
I
25
26
I
I
I
ities i returns are perfectly positively correlated, the
aggregation of many securities in a portfolio yields a total
risk position that is less than the simple summation of the
individual risks of each security (weighted by its importance
to the portfolio). That is, the risk of the portfolio is
less than the weighted average risks of its individual parts.
This concept of risk reduction is called diversification.
The power of the portfolio formulation is that, although
the risk of the portfolio is less than the sum of the risks
of each security held in the portfolio, weighted by their
relati ve importance to the portfol io, the expected return of
the portfolio is exactly equal to the summation of each
securi ty' s expected return, again weighted by their relative
importance to the portfolio. In other words, an investor
could form a well diversified portfolio in such a fashion as
to incur only a fraction of the security's undes irable fea-
tures--its risk--while obtaining all of the security's desir-
able features--expected return. Because of this, Markowitz
argued that rational, risk averse investors should construct
well diversified efficient portfolios.
Q. Is the Markowitz approach still used for the actual deriva-
tion of portfolios?
A. Yes. The Markowitz approach to portfolio formulation was
originally based on the separate interrelationships between
the returns on each security held in a portfolio. For
example, if there were four securi ties in a portfolio, when
TESTIMONY - 32
I
I
I 1
2
I 3
I
4
5
I 6
I 7
8
I 9
I 10
11
I 12
I 13
14
I 15
I 16
17
I 18
I 19
20
I 21
I 22
23
I 24
I 25
26
I
I
I
the Markowitz approach is used,. one must determine the sepa-
rate interrelationships between security 1 with security 2;
securi ty 1 with security 3; security 1 with security 4;
securi ty 2 with security 3; security 2 with security 4; and
security 3 with security 4. Markowitz noted, however, that
his original formulation could be vastly simplified if one
could identify a factor that was common to each security.
Rather than having to compute the separate interrelationships
between each of the securities in a portfolio, one could
simply estabish what the interrelationship was between each
individual security and some common factor. One could then
deduce from this simple interrelationship the implied sepa-
rate interrelationships of all the securities that were held
in the portfolio. Professor William Sharps, then of the
Univers i ty of Washington, constructed such a common factor in
his article, "A Simplified Model for Portfolio Analysis",
published in 1963. Sharpe's construction is called the market
model because the one common factor affecting all securities'
returns is taken to be general market movements.
Sharpe's market model is stated as equation (1) of Sche-
dule 8 of Exhibit 101. In this formula, "Rit" refers to the
possible rates of return that common stock in firm "i" could
earn during àny time period "tl.iiit IImt represents the possible
rates of return that could be earned by the IImarket portfolioll
..
during any time period Ii til , and IIEitll represents the possible
error terms in the relationship between the paired values of
TESTIMONY - 33
--"R' "it and "..R IImt .This model implies that a security's return
for any time period lit" can be explained by one factor--the
general market's returns. This explanation, however, is not
perfect; thus, the error term, "Bit". "Beta", or "B", measures
the responsiveness of security "i'" s returns to chang ing
market returns. For example, if a firm's common stock beta
equals 1. 0, this implies that, on the average, whenever the
market's returns increase or decrease by so many percentage
points, the firm's common stock market returns will change by
the same number of percentage points, and in the same direc-
tion. That is, if the market's return increases by one full
percentage point and its beta equals one, then the firm's
stock return, on the average, will increase by one full per-
centage point. If "B" is less than 1.0, then for every per-
centage point change in the market, the security, on the
average, will change by fewer percentage points. The oppo-
site conclusion is reached if "B" is equal to some value in
excess of 1.0. For these reasons, stocks whose betas are
less than 1.0 are typically referred to as defensive stocks,
and stocks whose betas are greater than 1.0 are typically
referred to as aggressive stocks.
Q. Is these any other way to interpret the market model?
A.Yes.The tildes over the variables, "Ri t", "Rmt" and -"Eit"
in equation (1) of Schedule 8 of Exhibit 101 signify that
these variables are random variables: they can take on a
wide variety of values for any time period "t". This is to
TESTIMONY - 34
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21 Q.
22 A.
23
24
25
26
say that their specific values are uncertain. Therefore,
Sharpe's market model can be interpreted as indicating that
the uncertainty (or volatility) of a security's return has
two components" "BiRmt" and "Ei t" . The first component is
called the market risk of an individual security, and the
second term is typically called the unique risk of a security.
The market risk component might be thought of as the volatil-
i ty of the security's return that is induced by general market
movements. This source of risk, from the investors' perspec-
tive, is inescapable because the investor has no control over
general market movements. The unique risk component might be
described as the volatil i ty of the security's return that is
induced by factors that are unique to the firm (e.g., a strike,
an unexpected profitable event, etc.). Although market risk
is unalterable, an investor may eliminate unique risk if he
holds many different securities in a portfolio. In this
case, the individual unique effects of each firm can be
expected to, on the average, offset one another. For this
reason, unique risk is said to be diversifiable; it can be
eliminated.
What is the significance of the "B" term in this model?
Because a firm's beta measures the responsiveness of its com-
mon stock return to the market's changing return, "B"
measures the risk that the firm's common stockholders face
relative to the risk of the market in general. Because
market risk is unavoidable (nondi vers if iable) by investors,
TESTIMONY - 35
I
I
I
1
2
I 3
I
4
5
I 6
I
7
8
I 9
I
10
11
I 12
I 13
14
I 15
I 16
17
I 18
I 19
20
I 21
I 22
23
I 24
I 25
26
I
I
I
while unique risk is diversifiable, "B" provides a measure of
the relevant risk of the firm's common stock relative to the
market's risk. If the firm's beta is greater than 1.0, its
risk is greater than the market's general risk and the firm's
stockholders would demand relatively more compensation for
bearing this risk.
Q. How does one estimate the beta term for a specific firm?
A. The Sharpe Market Model is a simple linear model. A firm's
"B" is est imated by regress ing its past common stock returns
against the market's returns. This regression will provide
the best historic estimate of the firm's beta value. (Statis-
tically, the beta estimate generated from this regression is
said to be B. L.U. E. --the best ,linear , unbiased, estimate.)
Q. Is the market model the capital asset pricing model?
A. No. Thus far we have discussed portfolio theory which indica-
tes that, in an efficient capital market, most rational and
risk averse investors should construct well diversified port-
fol ios because the unexpected or unique risk of a security
c an be el imina ted.
Capital market theorists, during the mid-1960's, took
the .implication of portfolio theory as fact and wondered:
Given that all rational and risk averse investors will hold
well diversified portfolios, what will be the resulting
equilibrium relationship in the capital market between a
securi ty' s risk and its expected return? The answer to this
question, under additional assumptions, has come to be known
TESTIMONY - 36
I
I
1
I 2
I 3
4
I 5
I 6
7
I 8
I 9
10
I 11
I 12
13
I 14
I 15
16
I 17
I 18
19
I 20
I 21
22
I 23
I 24
25
I 26
I
I
I
as the capital asset pricing model.
Q. Please explain the capital asset pricing model.
A. Symbolically, the capital asset pricing model can. be stated
as shown in equation (2) of Schedule 8 of Exhibit 101. In
this equation "E(Ri) II represents the rate of return that
stockholders in any firm (firm "i") should expect to receive
over some future period. It consists of the riskfree rate of
return that will actually prevail in the market throughout
the future period, "Rf", plus a risk premium. The risk pre-
mium for firm II i" is found by multiplying its beta that will
prevail throughout the future period, times the "market risk
-.premium", "( E (Rm) - (Rf)," that stockholders expect to pre-
vail in the economy during the future period. The market
risk premium measures the additional compensation, in excess
of the riskfree rate of return, that investors must expect to
receive for bearing the risk associated with investing in
common stock that is of average market risk. If a specific
firm's common stock is not as risky as the average market
risk, the inves tors should not demand to be compensa ted by
II (E(Rm) - Rfl"; they should receive less compensation.
Because "Bi" measures the risk of the firm's common stock
relative to the market, the multiplication of "Bi" times the
market risk premium results in the minimum risk premium--in
excess of the riskfree rate--necessary to compensate
investors for bearing the risk of common equity in firm "i".
Q. How can equation (2) of Schedule 8 of Exhibit 101, the CAPM
TESTIMONY - 37
I
I
I 1
2
I 3
I 4
5
I 6
I 7
8
I 9
I 10
11
I 12
I 13
14
I 15
I 16
17
I 18
I 19
20
I 21
I 22
23
I 24
I 25
26
I
I
I
method, be used to determine Intermountain's cost of common
equity?
A. Equation (3) of Schedule 8 restates equation (2) specifically
in terms of Intermountain's common stock. Equation (3) states
that the minimally acceptable rate of return that Intermoun-
tain's stockholders must expect, before they will divert their
funds away from riskfree investing and into investments in
common stock that are alike in risk to Intermountain, can be
estimated by multiplying the most likely value of Intermoun-
tain's beta that is expected to prevail in the market during
the future period, times the market risk premium that is
expected to prevail over the future period; one then adds to
this amount the riskfree rate of return that will prevail
during the future period.
Q. Professor Campbell, I would like to have you discuss the esti-
mation of each of these three terms separately. First, have
you undertaken a study to determine the market risk premium
to be used to estimate Intermountain's cost of common equity?
A. Yes, I have.
Q. I would like to have you describe the issues that are involved
in the determination of the market risk premium, including
why we should expect a market risk premium to exist.
A. Investors have many investment media from which to choose.
U. S. Treasury securities represent one source of investing.
I f an inves tor currently purchased aU. S. Treasury securi ty,
then he could precisely predict what future cash flows would
TESTIMONY - 38
I
I
1
I 2
I 3
4
I 5
I 6
7
I 8
I 9
10
I 11
I 12
13
I 14
I 15
16
I 17
I 18
19
I 20
I 21
22
I 23
I 24
25
I 26
I
I
I
be available from this investment. For this reason u. S.
Treasury securities are called riskfree securities. Another
source of investing is in, say a mutual fund, consisting of
many different shares of common stock.If the investor
invests in the mutual fund, however, he realizes that the
future cash flows associated with this investment are uncer-
tain. Therefore, before the investor can be expected to
divert his investment away from riskfree U. S. Treasury secur-
i ties, into risky common stock, he must expect, on average,
to earn a greater rate of return than could have been
achieved by riskfree investing.
It is important to note that investors do not require
that the actual rate of return associated with risky securi-
ties always provide a return in excess of the riskfree rate;
but, the return associated with the risky investment must be
expected to be higher, on average, than the riskfree rate
that prevails over the same period. If the investor did not
expect to earn, on average, a return greater than that by
riskfree investing, he would not subject his capital to the
additional risk.
Q. How can an economist estimate the magnitude of the risk pre-
mium investors demand?
A. Because investors expect to receive a positive risk premium,
on average, economists attempt to estimate the "average" risk
premium that exists in the market. While this procedure is
simple in concept, the determination of such an "average"
TESTIMONY - 39
I
I
I 1
2
I 3
I 4
5
I 6
I 7
8
I 9
I 10
11
I 12
I 13
14
I 15
I 16
17
I 18
I 19
20
I 21
I 22
23
I 24
I 25
26
I
I
I
premium is beset with many diff icul ties. The most important
problem being the determination of what is "average".
For example, it would be very easy to measure the actual
return that was achieved by common stock investment over some
period of time, and subtract from that rate the return that
could have been achieved by investing in a riskfree security.
Following this procedure for several periods, adding the
results, and finding the arithmetric average would result in
a calculated "average market risk premium". The diff icul ty
with this procedure is the determination of whether the cal-
culated "average" value represents the average risk premium
demanded by investors: it may be much lo~ler than what was
demanded, or much higher. Indeed, one might find that the
calculated "average" risk premium over some periods was nega-
tive. It would be faulty logic, however, to conclude from
this finding that investors appear to demand a negative risk
premium. A more logical conclusion would be that the period
of time that was used to find the arithmetric average was
atypical.
Another important problem in the measurement of a risk
premium is def ini tional, not conceptual. For example, the
market risk premium that must be estimated in the capital
asset pric ing model is symbol ically def ined as: .. E (Rm - Rf)".
In other words: the expected market risk premium is the
additional expected compensation, in excess of the riskfree
return that could have been achieved, that investors demand
TESTIMONY - 40
I
I
1
I 2
I 3
4
I 5
I 6
7
I 8
I 9
I
10
11
I 12
I
13
14
I 15
I
16
17
I 18
I
19
20
I 21
I
22
23
I 24
I 25
26
I
I
I
before they will invest in the common stock contained in the
market. The difficulties with this definition are: what is
meant by the "market", and what is meant by a "riskfree" asset.
So, the three problems in this area concern the def ini tion of
the market, the definition of the riskfree rate and the
period of time one chooses to measure meaningful averages.
Two very common surrogates, widely used to measure the
market, are the S&P and the NYSE composite (both on a value-
weighted, with dividends basis). Because my beta calcula-
tions are derived from the NYSE composite listed in the CRSP
data files, I use this as the definition of the market port-
fol io and report these real i zed returns in Schedule 4 of
Exhibi t 101.
As I previously indicated, u.s. Treasury securities are
used as the riskfree asset, from wh ich one measures the
r iskfree rate of return.
Q. Please describe the CRSP data files that you have used.
A. The Center for Research in Security Prices (CRSP) data files
are maintained at the University of Chicago. (They were pre-
viously maintained by Standard and Poors.) This data file
contains monthly price and return information for all common
stocks listed on the New York Stock Exchange. Additionally,
this file contains the monthly returns associated with a
weighted by total value composite of all common stock listed
on the New York Stock Exchange. Because this composite return
is generally taken to be the market portfolio, the CRSP file
TESTIMONY - 41
I
I
1
I 2
I 3
4
I 5
I 6
7
I 8
I 9
10
I 11
I 12
13
I 14
I 15
16
I 17
I 18
19
I 20
I 21
22
I 23
I 24
25
I 26
I
I
I
is particularly useful for CAPM studies; it is widely used by
the academic and business communi ties. Because of its use-
fullness many universities and investment firms subscribe to
this data.
Q. Which U. S. Treasury Security do you use to measure the market
risk premium?
A. I have used short-term U. S. Treasury bill rates--30 day rates
and 9l-day rates.
9l-day Treasury bills yields are reported in the second
edition (1977) of A History of Interest Rates, by Sidney Homer
(pp. 374-375) published by Rutgers University Press. These
data, stated on an annual coupon basis, have been updated
through the tenth month of 1979 using data contained in the
Federal Reserve Bulletins, and are presented in Schedule 5 of
Exhibit 101.
30-day Treasury bill yields are reported in Ibbotson and
S inquef ield' s study for the Financial Analysts Research
Foundation (1977) entitled Stocks, Bonds, Bills and Inflation:
The Past and the Future.These data have been updated
through December 1979 by Compater Directions Advisors, Inc.,
of Silver Springs, Maryland, and they are presented in
Schedule 6 of Exhibit 101.
Using the data in Schedules 4, 5 and 6 enables one to
-estimate the annual average market risk premium, "(E(Rm) - RfJ",
used in the capital asset pricing model.
Q. Dr. Campbell, what period of time did you use to determine
TESTIMONY - 42
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
1
2 A.
3 Q.
4 A.
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
the average market risk premium?
I used the period starting with 1926 and extending through 1979.
Why did you analyze this period?
Ideally, if the data were available, I would have analyzed
the period starting with 1920. But, the CRSP data files, on
which I relied to determine the past returns on the market
portfolio, do not report data prior to 1926.
As I previously indicated, if one is attempting to
measure the "average" market risk premium, one must be very
sensitive to the time period that is used. For example, due
to short cycles of activity, one might calculate a return
over some short time period that results in an abnormally
high average risk premium, or the converse. Therefore, the
duration chosen to measure the market risk premium must be
long enough to average out the potentially atypical sub-
periods. Moreover, the time period chosen should contain a
complete cycle of activity, by which I mean a period of time
that begins with a peak of activity, moves through a trough,
and then reaches another peak. The reason why a full cycle
is important is because potential distortions caused by the
first half-cycle of downward movements can be expected to be
offset by opposite distortions caused by the second half-
cycle of upward movements.
As an illustration of the necessity of including a com-
plete cycle of activity, suppose that we are interested in
the average time it takes an individual to run two miles.
TESTIMONY - 43
I
I
1
I 2
I 3
4
I 5
I 6
7
I 8
I 9
10
I 11
I 12
13
I 14
I 15
16
I 17
I 18
19
I 20
I 21
22
I 23
I 24
25
I 26
I
I
I
TESTIMONY - 44
I
I
I 1
2
I 3
I 4
5
I 6
I 7
8
I 9
I 10
11
I 12
I 13
14
I 15
I 16
17
I 18
I 19
20
I 21
I 22
23
I 24
I 25
26
I
I
I
with 1920 and extends at least to the present.
In 1920 a new peak of both long-term and short-term
interest rates was reached. Thereafter these yields secu-
larly decreased, reaching a low in the 1940' s. This secular
decrease can be noted from the U.S. Treasury bill data in
Schedules 5 and 6 of Exhibit 101. In both cases, the yields
decreased to roughly 1% in 1931 and 1932 and thereafter
reached a low in 1940. In the subsequent nine years however,
the rates remained extremely low, staying well below 1%.
Because of these secularly decreasing and low rates of
interest, the first half-cycle will be def ined as 1926-1949.
Just as in the analogy of telling the runners who have
reached the bottom of the hill to turn around and run back up
the hill in order to ensure that we have a complete cycle, we
want to analyze the second half-cycle as that period of time
that reverses the first half-cycle. This means that we want
to select as the second half-cycle a period where interest
rates start low and gradually get larger.(The reverse of
the first half-cycle of initially high interest rates and a
gradual decay.) Moreover, if there are any other events that
may have profoundly influenced the first half-cycle (on the
way down), we would like to include those events in the
second hal f-cycle (on the way up).
As Schedules 5 and 6 portray, it is doubtful that the
second half-cycle has been completed as of yet: the final
peak may not have been reached. Moreover, because this series
TESTIMONY - 45
I
I
1
I 2
I 3
4
I 5
I 6
7
I 8
I 9
10
I 11
I 12
I
13
14
I 15
I
16
17
I 18
I
19
20
I 21
I
22
23
I 24
I
25
26
I
I
I
of data begins (1926) six years after the original peak
( 1920), to complete a full cycle we would want to choose that
period of time that corresponds to five or six years after
the new peak is reached;. whenever that is. Nonetheless,
there are sufficient symmentries in the financial data from
1926-1949 and 1950-1979 to warrant the study of this period
in order to find meaningful "average" results.
For example, in the first half-cycle (1926-1949), the
highest annual yield (reached in 1929) was about four times
the average yield for the period. In the second half-cycle
(1950-1979), the highest yield (1979) was roughly three times
the average yield for this period. In both cases the high
yields occurred at app~oximately reverse times--near the
beginning of the first half-cycle, and near the end of the
second half-cycle. Additionally, near the beg inning of the
first half-cycle there was a great depression in the economy
( 1929-1931). And near the end of the second half-cycle, the
second greatest depression (as measured by losses on the
stock market) in this century occurred (1973-1974). Moreover,
to the extent that wars cause potential distortions to the
economy, World War II occurred during the first half-cycle,
and the Korean, Vietnam and the Cold War occurred in the
second half-cycle. Finally, let me note that at the beginning
of the first half-cycle, short-term rates were greater than
long-term rates and currently, short-terms rates are again
greater than long-term rates. Thus, these two half-cycles
TESTIMONY - 46
I
I
I 1
2
I 3
I 4
5
I 6
I 7
8
I 9
I 10
11
I 12
I 13
14
I 15
I 16
17
I 18
I 19
20
I 21
I 22
23
I 24
I 25
26
I
I
I
contain many opposite financial parallels.
Overall, yields on short and long-term financial securi-
ties from 1926-1979 can be pictured as the inside of a bowl;
1926-1940 corresponds to downward movements to the bottom of
the bowl; 1940-1949 as movements along the bottom of the
bowl; and 1950-1979 represents movements back up the other
s ide of the bowl. Again, it is not clear if we have moved to
the outer lip of the bowl, but it is clear that its height
has already surpassed the previous peak. In any event, due
to the reverse symmetries, one can expect countervailing
biases over this period to roughly cancel out.
Q. What biases are you speaking of?
A. Consider investors in 1920. Interest rates were at their
peaks and inflation was relatively high. Over the future
period inflation and interest rates secularly decreased.
Suppose investors, at the beginning of this period, priced
the market portfolio to yield, on an expected basis, 8% more
than u.s. Treasury securities. But, because inflation was
decreasing and the yields on U. S Treasury securi ties were
decreasing, then the price that they would be willing to pay
for the market portfolio, while still expecting to receive an
8% differential in the future, must have increased through
this period. And, with these increasing prices at lower and
lower expected rates of future inflation, the original inves-
tors might have found that, on average, they actually received
a rate of return somewhat greater than they had expected to
TESTIMONY - 47
I
I
I 1
2
I 3
I 4
5
I 6
I 7
8
I 9
I 10
11
I 12
I 13
14
I 15
I 16
17
I 18
I 19
20
I 21
I 22
23
I 24
I 25
26
I
I
I
recei ve.
Now assume the reverse. Investors are assumed to start
at 1950. Throughout the next 30 years inflation and yields
on U. S. Treasury securities were secularly increasing. The
first investors in 1950 price the market portfolio to achieve,
on an expected bas is, say 8% more, on average, than the yields
on U. S. Treasury securities. But as we progress through the
period we find that new investors, forecasting even higher
rates of future inflation, are willing to pay less and less
for the market portfolio in order to expect to receive, on
average, an 8% risk premium. but, associated with falling
market prices is the fact that the original investors watch
the value of their market portfolios decrease to perhaps small
positive risk premiums or perhaps even negative risk premiums.
These represent the potential biases present in using
actual risk premiums as surrogates for the investors' demanded
(expected) risk premium. Fortunately, however, the series of
data under analysis does not strictly conform to falling and
rising inflation or yields on short and long-term instruments.
For example, looking at the second half-cycle, 1950-1979 in
Schedule 5 of Exhibit 101, one can identify many sub-cycles
of rising and falling economic activity. For example, the
period 1953-1957 represents one short but complete cycle of
peak, trough and peak. 1957-1959 represents an even shorter
sub-cycle in this period. 1959-1966, 1966-1969, 1969-1974
and 1974-1979 are examples of other sub-cycles within this
TESTIMONY - 48
I
I
I
1
2
I 3
I
4
5
I 6
I
7
8
I 9
I
10
11
I 12
I
13
14
I 15
I 16
17
I 18
I 19
20
I 21
I 22
23
I 24
I 25
26
I
I
I
second half-cycle. Thus, during the second half-cycle (1950-
1979), although inflation and Treasury yields were rising on
average, there were probably suff icient shorter cycles of
activity that removed a good deal of the potential bias that
could have been created during this period.
Q. Dr. Campbell, what were your findings for this study?
A. Over the first half-cycle of economic activity (1926-1949),
spanning 24 years, the average realized return on the market
portfolio, in excess of the 30-day u. S~ Treasury yields, was
8.41 %.
Over this same period, the average realized return on
the market portfolio, in excess of 9l-day U. S. Treasury
yields, was 8.32%.
Over the second half-cycle of economic activity (1950-
1979), spanning 30 years, the average realized return on the
market portfolio, in excess of the 30-day U. S. Treasury
yields, was 8.37%.
And, over this same period, the average realized return
on the market portfolio, in excess of the 9l-day U. S. Treasury
yields, was 8.11%.
It is interesting to note that the realized risk premium
for both 30-day and 9l-day U. S. Treasury securities was
greater (although slight) in the first half-cycle of activity
relative to the second. As previously explained, this was to
be expected because of the potential biases involved in eval-
uating periods characterized by secularly decreasing and
TESTIMONY - 49
I
I
I 1
2
I 3
I 4
5
I 6
I 7
8
I 9
I 10
11
I 12
I 13
14
I 15
I 16
17
I 18
I 19
20
I 21
I 22
23
I 24
I 25
26
I
I
I
increasing inflation rates.
Q. Dr. Campbell, what do you conclude from this study?
A. The average market risk premium, over both half-cycles, 1929-
1976, was about 8.4% using 30-day Treasury bill data as the
r iskfree surrogate. using 9l-day Treasury bill yields, the
average market risk premium over this time was 8.2%. The
average of these two market risk premiums, 8.3% represents a
reasonable val ue to be used in these proceed ings. I would
like to add however, that if data were available from 1920,
and as we received new data corresponding to periods that
lead up to the new peak of interest yields that signifies an
end in the cycle, I believe that one would find that the
average risk premium equalled 9.0%. And, I believe that this
value is more representative of what investors, on average,
expect. The market risk premium is stated on a before trans-
action cost basis: before brokerage commissions and before
personal taxation. Adjusting for these costs probably results
in a 4 1/2% risk premium, not 9%, that investors can actually
expect to earn. And a 4 1/2% expected risk premium, relative
to riskfree investing, does not strike me as excessively high
at all. Indeed, I arrive at the reverse conclusion.
Q. Dr. Campbell, wiii you summarize why it is appropriate to use,
as an estimate of the future market risk premium that inves-
tors expect to achieve, the market risk premium that was
achieved on average, in the long run?
A. If short time durations were used, the time period selected
TESTIMONY - 50
I
I
I
1
2
I 3
I
4
5
I 6
I 7
8
I 9
I 10
11
I 12
I 13
14
I 15
I 16
17
I 18
I 19
20
I 21
I 22
23
I 24
I 25
26
I
I
I
might be dominated by recessions on short peaks of economic
activity. Using short periods of time as indicative of what
investors were expecting could be very misleading because one
would be implicitly assuming that the bust or boom period
that characterized this short time period would prevail
indef ini tely.
More specifically, the stock market's returns are very
volatile. But this fact should not be surprising. Indeed,
to say that common stock sec uri ties are risky is really to
say that their prices will fluctuate. Because the stock mar-
ket serves as the center for risk bearing, volatil i ty should
be expected. Small changes in economic factors will be mag-
nified in the process of bearing this new perceived risk.
But, due to the volatility of the market, and the volatility
of the resulting risk premiums actually earned, there is only
one statistic of interest--its. long run average value. The
best prediction of next year's actual market risk premium is
its past long run value, 8.3%. At times, such as last year
(1979), this prediction will be very close to what actually
transpired because the actual market premium last year was
8.5% using 30-day Treausry data, and 8% using 9l-day data.
At other times, such as 1977, the actual market risk premium
may be much lower than expected (-10 %); and at other times,
such as 1976, the actual market risk premium may be much
higher than expected (21 %) . Nonetheless, statistically, our
very best estimate for next year's market risk premium is 8.3%.
TESTIMONY - 51
I
I
I 1
2
I 3
I 4
5
I 6
I 7
8
I 9
I 10
11
I 12
I 13
14
I 15
I 16
17
I 18
I 19
20
I 21
I 22
23
I 24
I 25
26
I
I
I
Q. Are there any other studies that you are aware of that sup-
ports your estimate of an 8.3% market risk premium?
A. Yes. Ibottson and Sinquefield published a widely read
article in the Journal of Business (January 1976) pp. 11-47,
entitled: "Stocks, Bonds, Bills and Inflation: Year by Year
Historical Returns (1926-1974)". Using 30-day Treasury rates
and def ining the market portfol io as the S&P 500, these
authors show that the average market risk premium is 8.8%.
Using more recent data (1950-1974), Ibottson-Sinquef iled' s
data show an 8.2% market risk premium.
Thus, I am rather confident in my estimate of a required
market risk premium of 8.3%.
Q. Have you undertaken a s.tudy to determine Intermountain's beta?
A. Yes. Intermountain's stock price and dividend data is
reported in Standard and Poor's ISL: Daily Stock Price Index
(O-T-C). Using these data, which are assembled in Schedule 7
of my Exhibit, I calculated Intermountain's monthly rates of
return. The market portfol io' s returns for these periods
were extracted from the CRSP data files and are also con-
tained in Schedule 7. Regressing Intermountain's monthly
returns against the market's returns, Intermountain's historic
seven-year beta was found to equal .72, with a standard error
of .13; its his tor ic six-year beta equalled .69, with a stan-
dard error of .15; and its historic five-year beta equalled
. 84, with a standard error of .19. These data imply that, if
anything, Intermountain appears to becoming more risky--its
TESTIMONY - 52
I
I
I 1
2
I 3
I 4
5
I 6
I 7
8
I 9
I 10
11
I 12
I 13
14
I 15
I 16
17
I 18
I 19
20
I 21
I 22
23
I 24
I 25
26
I
I
I
stock price movements is following the market's general move-
ment more closely.
Q. Which of these historic betas do you use for Intermountain?
A. Past research in this area indicates that one should analyze
a firm's beta over a period of five to seven years.
For example, Professor Fama of the Uni vers i ty of Chicago
states in his Foundations of Finance (Basic Books, 1976) on
page 132 that: "With monthly data the optimal estimation
period is apparently five to seven years." And, Professors
Tinic and West, in their Investing in Securities: An
Efficient Market Approach (Addison-Wesley Publishing Co.,
1979) on page 179, not that in order to calculate a historic
beta that a
preliminary step involves the choice of the histor-
ical period over which to run (the) regression. As
with all forecasting techniques based on extrapo-
lating past behavior (to the future), there is a
dilemma between using a long series, which-it is
hoped-give better statistical results, and using a
shorter, more current series, which avoids using"stale" data as the bas is for proj ections. While
there is no clear-cut answer to this dilemma, pre-
sent empirical studies have shown that predictions
of betas based on regressions involving five toseven years of monthly returns ,,¡ere superior topr.edictions resulting from regressions using either
longer or shorter series. (My parentheticals.)
In Intermountain's case the average of its seven-year,
six-year and five-year betas is .75.
Q. Have you performed any other studies to support the reason-
ableness of a beta value of .75 for Intermountain?
A. Yes. I have evaluated a sample of gas distribution companies
TESTIMONY - 53
I
I
I 1
2
I 3
I 4
5
I 6
I 7
8
I 9
I 10
11
I 12
I 13
14
I 15
I 16
17
I 18
I 19
20
I 21
I 22
I
23
24
I 25
I
26
I
I
listed in April 18, 1980 's Value Line that are somewhat com-
parable to Intermountain. These firms and their reported
values of beta follow: Alagasco, Inc. (.55); Atlanta Gas
Light (.60); Bay State Gas (.60); Brooklyn Union Gas (.50);
Cascade Natural Gas (.60); Gas Serv ice Co. (.60); Ind iana Gas
( .55); Laclede Gas (.65); Northwest Natural Gas (.80);
Wiscons in Gas (.55).
Relati ve to these gas distribution companies, Intermoun-
tain is relatively more risky except for Northwest Natural Gas.
Q. Why do you suppose that Intermountain and Northwest Natural
are riskier than the other firms in this group?
A. I believe the difference can be accounted for by two separate
reasons:(1) the compos i tion of revenue by customer-class
and (2) their capital structure.
In terms of the composition by revenue, Intermoun-
tain's residential revenue comprises only about 25% of its
total; 50% of its revenue is generated from more volatile
industrial customers. Northwest Natural Gas' residential
revenue is also very low (28%), with 56% of its revenue being
tied to volatile industrial demand. The other one firm in
this sample that joins Intermountain and Northwest Natural's
heavy reI iance on industrial revenue is Cascade Natural Gas
(56% industrial, 16% residential).
The other firms in this sample have a much higher com-
position of revenue generated from more stable residential
demand: Algasco 40%, Bay State Gas 53%, Brooklyn Union 81%,
TESTIMONY - 54
Indiana 45%, Laclede 58% and Wisconsin 43%.
In terms of the capital structure of these firms, Inter-
mountain's capital structure from 1975 through 1979 is
characterized by low common equity ratios (33% to 41%). In
other words, Intermountain has a lot of financial leverage
which acts to increase its market volatil i ty. Northwest
Natural Gas and Cascade Natural Gas, which share the same
revenue volatility induced by low concentration of residen-
tial sales, also have low equity ratios: Northwest Natural
(35% to 45%), Cascade Natural Gas (25% to 33%).
Q. Are there other financial similari ties between Natural,
Intermountain and Cascade?
A. Yes. I have evaluated the volatility of these firm's earn-
ings per share over the past five years (1975-1979). Measur-
ing the volatil i ty by the standard deviation, and forming the
ratio of the calculated standard deviation to the mean of its
earnings per share over the last five years, I found that the
firms \vi th the most volatil i ty (risk) of past earnings per
share, per uni t mean, were:Intermountain (.43), Cascade
(.43), Indiana (.32), Gas Service Co. (.29), Northwest
Natural (.27), Wisconsin (.25), Atlanta Gas Light (.23),
Laclede (.21), Brooklyn Union (.18), Bay State Gas (.15) and
Alagasco (.14).
Q. Did you evaluate the bond ratings of these firms to discern
how risky they were viewed from the perspective of financial
rating agencies?
TESTIMONY - 55
I
I
I 1
2
I 3
I 4
5
I 6
I 7
8
I 9
I 10
11
I 12
I 13
14
I 15
I 16
17
I 18
I 19
20
I 21
I 22
23
I 24
I 25
26
I
I
I
A. Yes. Al though Intermountain is not rated, the Company
believes that is would be rated BAA. After reviewing
Intermountain's total interest coverage ratios, however, I
feel that a BAA is the maximum bond rating the Company could
expect; a BA rating would be more likely.
The highest rated (safest) firm in this sample was
Laclede (Aa). Its total interest coverage, as reported in
Value Line for 1979 was 4.6.
The next safest firms were Indiana Gas (A, coverage
ratio 6.1), Wisconsin Gas (A, coverage ratio 4.9), Northwest
Natural (A, coverage ratio 3.6), Brooklyn Union (A, coverage
ratio 3.5), Gas Service Co. (A, coverage ratio 3.2) and
Atlanta Gas (A, coverage ratio 3.0).
The most risky firms were Alagasco (A to BAA, coverage
ratio 2.9), Bay State Gas (BAA, coverage ratio 2.3) and
Cascade tJatural Gas (B-BA, coverage 3. 0) .
Although Intermountain has a recent total coverage ratio
of 3.2, prior to this year it was but 1.5. Therefore, unless
the rating agencies felt that Intermountain could maintain
this current coverage ratio, its likely that they would be
rated BA.
Q. What conclusion have you arrived at based on this study of
different gas distribution companies?
A. In terms of market risk, lJorthwest tJatural and Intermountain
are the most risky.
In terms of having to rely on i~re volatile customer
TESTIMONY - 56
demand, Cascade Natural Gas, Intermountain and Northwest
Natural have the greatest proportion of revenue generated
from more volatile industrial and commercial demand.
In terms of the riskiest capital structure over the past
five years, Cascade Natural Gas had an average common equity
ratio of 29.0 and Intermountain was the next riskiest (36.6).
Indiana Gas (36.8), Brooklyn Union (37.5), Atlanta Gas Light
(37.8), Bay State Gas (38.2) and Northwest Natural (39.4)
have very similar capital structures. The most secure capi-
tal structures are enjoyed by Gas Serv ice Co. (40.0),
Wisconsin Gas (43.9) and Alagasco (48.8).
In terms of measuring the volatility of past five-years
earnings, divided by the mean of its earnings, intermountain
and Cascade have the highest risk ratios of .43. The next
three firms of highest risk per unit mean were: Indiana, Gas
Service Co. and Northwest Natural.
In terms of bond ratings, the riskiest firms were
Cascade Natural Gas, Intermountain, Bay State Gas and Alagasco.
The conclusion I have formed is that Intermountain,
Northwest Natural and Cascade Natural Gas are probably the
three most comparable firms in this group. Intermountain and
Northwest Natural have about the same market risk and Inter-
mountain and Cascade Natural share many of the same financial
and reg ional characteristics.
I must say that what troubles me most is when a firm
1 ike Brooklyn Union is evaluated as a comparable firm.
TESTIMONY - 57
I
I
I 1
2
I 3
I
4
5
I 6
I
7
8
I 9
I
10
11
I 12
I 13
14
I 15
I 16
17
I 18
I 19
20
I 21
I 22
23
I 24
I 25
26
I
I
I
Brooklyn Union competes in a New York market where electri-
city costs four to six times the amount paid in the Northwest.
Qui te frankly, due in large part to the abundance of cheap
hydro power in the Northwest, and the more recent gl ut of
residual oil from Alaska, natural gas companies in the North-
west have had a difficult time competing like gas companies
on the east coast. These factors have led to a great deal of
volatility in their earnings. Rather than having a high com-
posi tion of residential demand to dampen this volatility, the
natural gas companies in the Northwest have seen their earn-
ings respond radically to weather changes and to industrial
demand changes. In other words, the greatest proportion of
earnings volatility is induced by factors over which manage-
ment has no control. Hence, the increased need for rate
relief.
Based on this study, I also conclude that a historical
beta val ue for Intermountain of .75 is reasonable. In this
spectrum of somewhat comparable firms, Intermountain is
riskier than the average but less risky than the most volatile.
Q. Is the val ue of the historic beta of a company's common stock
the value that should be used to estimate its stockholders'
minimally acceptable rate of return according to CAPM?
A. No. The appropriate value of beta is the value that stock-
holders expect to prevail during the future period. And, as
a general rule, historically generated betas have to be
adj usted to better estimate the future.
TESTIMONY - 58
I
I
I 1
2
I 3
I 4
5
I 6
I 7
8
I 9
I 10
11
I 12
I 13
14
I 15
I 16
17
I 18
I 19
I
20
21
I 22
I 23
24
I 25
I 26
I
I
Q. Why do you have the opinion that predictabili ty can be
improved if historic betas are adj usted?
A. Professor Marshall Blume published two widely read papers in
The Journal of Finance in March of 1971 and June of 1975 ("On
the Assessment of Risk", and "Betas and Their Regression Ten-
dencies"). Blume empirically verified what many researchers
had long suspected: historical betas have a natural tendency
to regress towards the value 1.0. That is, if a security or
portfolio had a historic beta of less than 1.0, there was a
greater than 50% probability that its future beta would be
greater than its past beta. Alternatively, if a security or
portfolio had a historic beta greater than 1.0, there was a
greater than 50% probability that its future beta would be
less than its past beta.
Q. How does one explain this natural tendency for betas to
migrate towards the value 1.0?
A. One interpretation of this evidence is that firms do not like
extreme positions. For example, a firm whose operations or
capi tal structure makes its common stock rather risky is
1 ikely to attempt to move back to a more moderate risk posi-
tion. And, a firm whose operations and capital structure
makes its common stock returns very predictable, but rela-
tively unprofitable, is likely to move into a more aggressive
posi tion, either operationally or financially, in an attempt
to increase its common stock returns. Thus, the empirical
fact that firm's betas have a natural tendency to migrate
TESTIMONY - 59
I
I
I 1
2
I 3
I 4
5
I 6
I 7
8
I 9
I 10
11
I 12
I 13
14
I 15
I 16
17
I 18
I 19
20
I 21
I 22
23
I 24
I 25
26
I
I
I
towards the value of one may not be too surprising.
Q. Did Professor Blume indicate that historic betas should be
adjusted to incorporate this migration tendency in order to
improve its future predictabili ty?
A. Yes. Dr. Blume showed that historic betas should be adjusted
in order to more closely approximate the future period beta
value.
Q. Have any other researchers shown that historic betas should
be adjusted to more closely approximate the future period
beta value?
A. Yes. This is a well accepted fact in the field of finance.
Two additional articles in The Journal of Finance ("The
Adjustment of Beta Forecasts," 30 (1975) pp. 1123-1128 by
Professors Klemkosky and Mart in; and "Are Betas Best," 33
(1978) pp. 1375-1984 by professors Elton, Gruber and Urich)
are examples of additional research that unequivocally
conclude that historic betas should be adjusted in order to
more closely predict the future period beta value. Moreover
Merrill-Lynch, Pierce, Fenner and Smith as well as Value Line
also adj ust the betas they report.
Thus, because we are trying to predict Intermountain's
future beta, we must adjust our historic estimate of .75 to
increase the accuracy of the estimate.
Q. Professor Campbell, what methods can one use to adjust the
historic value of beta to improve its future predictability?
A. Three popular adjusting methods are the Blume technique, the
TESTIMONY - 60
I
I
I 1
2
I 3
I 4
5
I 6
I 7
8
I 9
I 10
11
I 12
I 13
14
I 15
I 16
17
I 18
I 19
20
I 21
I 22
23
I 24
I 25
26
I
I
I
Vasicek (or Bayesian) method ("A Note on Using Cross-Sectional
Information in Bayesian Estimation of Security Betas, II The
Journal of Finance, 28 (1973) pp. 1233-1239), and the Merrill-
Lynch method.
Of these three methods the Merrill-Lynch (Pierce, Fenner
and Smith) method is the most straightforward. In order to
improve the predictability of future betas, historic betas are
adj us ted by multiplying .65 times the historic value and adding
to this .35 times the value 1.0. As is readily apparent, by
giving some weight to the value 1.0 (our best estimate of the
company's beta if we had no additional information), this
lowers (raises) the predicted beta relative to the historic
value if its historic value was greater (smaller) than 1.0.
Again, this technique incorporates the fact that betas have a
natural tendency to migrate to the value 1. O.(This adj ust-
ment technique is explained in Professor Sharpe's book:
Investments (1978) Prentice Hall, pp. 274-277.)
Applying the Merrill-Lynch technique to Intermountain's
historic value of .75 implies that the best etimate of its
future beta is .84; ((.75)(.65) + (.35)(1.0)). This value
is almost exactly equal to Intermountain's more recent five-
year historical beta of .85.
Q. Dr. Campbell, would it be appropriate to use the historical
calculated beta as the lower range and adjusted beta as an
upper range to reach an opinion concerning Intermountain's
future beta?
TESTIMONY - 61
I
I
I 1
2
I 3
I 4
5
I 6
I 7
8
I 9
I 10
11
I 12
I 13
14
I i5
I 16
17
I 18
I 19
I
20
21
I 22
I
23
24
I 25
I
26
I
I
A. It would be incorrect to use an unadj usted historical beta
value to estimate the future value of beta. All studies in
this area unequivocally show that past historical betas
should be adj usted to take into account their natural migra-
tion tendencies. The adjusted betas improve their predicta-
bility; the adjusted beta represents the best estimate of the
future beta. It would be inappropriate to form a range of
beta val ues us ing, as the downward portion of the range, the
histor ic val ue of the firm's beta because of its inabil i ty to
predict future betas relative to the adjusted beta.
Q. Professor Campbell, what conclusions have you drawn based on
your market risk premium study and your beta study?
A. Based on the data I analyzed from 1926 through 1978, the mar-
ket risk premium, using 30-day u. S. Treasury bills as the
riskfree surrogate, was found to equal 8.4%. Over the same
time period, but using 9l-day u.s. Treasury bills as the
riskfree surrogate, it was found to equal 8.2%. If one prop-
erly incl uded the time period starting with 1920, I would
expect that the findings wuld have been closer to 9%, but
given that I do not have data for these earl ier periods, the
mid-point of 8.3% probably represents a conservative (poten-
tially downward biased) estimate of what investors expect to
receive in excess of the riskfree rate for a security of
average market risk.
Intermountain's common stock, however, is not as risky
as the average; it would be inappropriate to assume that
TESTIMONY - 62
I
I
I 1
2
I 3
I 4
5
I 6
I 7
8
I 9
I 10
11
I 12
I 13
14
I 15
I 16
17
I 18
I 19
20
I 21
I 22
23
I 24
I 25
26
I
I
I
Intermountain's investors would require 8.3 percentage points
more than the riskfree rate that could be earned by riskfree
investing. The appropriate risk premium for Intermountain in
excess of the riskfree rate of return, is 7.0%; ((.84)(8.3%)).
In order to determine Intermountain's stockholders'
minimally acceptable rate of return, this minimally required
risk premium must be added to the near-term riskfree rate of
return that investors expect to achieve.
Q. Professor Campbell, what riskfree rates of return can inves-
tors currently earn?
A. Using quotations from the Wall Street Journal, on April 18,
1980, and May 5, 1980, during the last month an investor
could have purchased 3Q-day Treasury bills that yielded a
10.2% to 16.1% rate of return (stated on an annualized com-
pounded bas is) .
Three month U. S. Treasury bills purchased during this
time period have yielded 11.7% to 15.2% rate of return.
Six-month U. S. Treasury bills have yielded a 10.6% to
14.7% return.
One-year maturing U. S. Treasury notes could have been
purchased and, with certainty, the investor would have
received alD. 2% to 13.0% rate of return over the next one-
year period.
Two-year U. S. Treasury notes have yielded 10% to 12%.
These rates could have been earned wi th certainty over the
next two-year period.
TESTIMONY - 63
I
I
I 1
2
I 3
I 4
5
I 6
I 7
8
I 9
I 10
11
I 12
I 13
14
I 15
I 16
17
I 18
I 19
20
I 21
I 22
23
I 24
I 25
26
I
I
I
Indeed, seven, eight and nine-year maturing U. S. Treasury
notes could have guaranteed a 10% to 11% rate of return, per
year, over the next consecutive seven, eight or nine years.
These high rates of return are not unique to u.s. Treasury
securi ties: all minimally acceptable rates of return on all
forms of financial securi ties--including bonds and common
s tock--are at unprecedented historically high levels. For
example, in terms of new bonds issued by utility companies:
Pacific Telephone, an ATT subsidiary with a bond rating of A,
recently sold new bonds at a coupon rate of 15.5%. Public
Serv ice of Colorado sold its A-rated bonds at a coupon rate
of 15%. Duke Power and Light sold its A-rated bonds at a
coupon rate of 14.875%. And Florida Light and Power recently
sold its A-rated bonds at 15.25 %.
I would like to also point out that over the past month
these rates have dramatically decreased from their previous
yields. I do not expect a lot of improvement in the future,
however. In the months of March and April, short-term
Treasury securities were yielding 16% to 19% and longer term
notes were yielding 13 1/2% to 15 1/2%. But these rates were
v iewed as be ing a temporary reflect ion of the Federal Reserve
Board's aggress i ve pol icy to decrease consumer demand for
goods and services. The rates that currently exist, long-
term rates of about 10% are also viewed as a temporary reac-
tion to the Fed's policy. Most analysts expect these rates
to increase over the near term when the Fed starts to ease
TESTIMONY - 64
I
I
I 1
2
I 3
I 4
5
I 6
I 7
I
8
9
I 10
I
11
12
I 13
I
14
15
I 16
I
17
18
I 19
I 20
21
I 22
I 23
24
I 25
I 26
I
I
its restraints. The conclusion that one draws from these
violent swings in yields is that the Fed over-reacted and
will have to moderate its position shortly.
Q. Is it your opinion that .these rates that you have reported
will continue at these levels over the near term?
A. Yes. I believe that the current longer term yields of 10% -
13.5% are representative of what average short-term Treasury
securi ties will yield over the next two to three years. But
I expect a great deal of volatil i ty in these short-term rates
relative to long-term rates. In large part, this opinion is
formed under what I regard to be the most conservative of
estimates for future inflation over the next three years
(14 1/2%, 10% and 9%). _ Moreover, this opinion is formed on
what I expect to occur in our capital markets over the next
two to three years.
Salomon Borthers' 1980 Prospects for Financial Markets
states that inflation, already at historically high levels,
is likely to accelerate through 1980. Credit demands by
businesses to finance its inflated inventory requirements, to
finance its inflated investment plans, and to maintain suf-
f icient working capital are expected to increase throughout
1980. In addition to this increased demand for funds by the
business sector is the fact the U. S. Treasury will have to
raise mammoth amounts of new short-term securities just to
finance (roll-over) its maturing debt. Additionally, the
federal deficit for fiscal 1980 is projected to equal fifty
TESTIMONY - 65
I
I
I
1
2
I 3
I
4
5
I 6
I 7
8
I 9
I 10
11
I 12
I
13
14
I 15
I 16
17
I 18
I 19
20
I 21
I 22
23
I 24
I 25
26
I
I
I
to ninety billion dollars depending on what forecast one reads.
The Federal Reserve System must raise this amount as well.
So, on the one hand, our capital markets will face
increasing financing needs by the business and government
sectors throughout 1982. On the other hand, investors have
decreased funds to meet these needs. Wage guidel ines of last
year and this year (7 1/2%, and 7 1/2% to 9 1/2% respectively),
concurrent with a 13.3% inflationary rate last year and per-
haps a 14 1/2% to 16% rate this year, have decreased and will
decrease even further investor supplied disposable income.
All of these factors--an increase in the demand for funds, an
unprecendented raging rate of inflation, and a decraease in
the supply of funds--c~rtainly limi t the amount by which
interest rates could fall over the next three years.
Q. Why do you have the opinion that inflation over the next
three years is expected to 14 1/2%, 10 % and 9%.
A. So far in 1980, we have had three consecutive quarteres of an
18+% increase in the Consumer Price Index. I am beginning to
think that my estimate of an average rate for 1980 of 14 1/2%
is really too low based on these facts. Even though infla-
tion may let up in forthcoming months, it must drop to very
low levels for an average of 14 1/2% to obtain. Probably a
closer estimate to this year's (1980) rate of inflation is
15% to 16%.
In order to check the consistency of my estimates, I have
looked at Data Resources, Incorporated's latest forecast,
TESTIMONY - 66
I
I
I 1
2
I 3
I 4
5
I 6
I 7
8
I 9
I 10
11
I 12
I 13
14
I 15
I 16
17
I 18
I 19
20
I 21
I 22
23
I 24
I 25
26
I
I
I
April 1980. DRI forecasts with a 50% probability that the
rates of inflation for 1980, 1981 and 1982 will be 15.3%,
11.6% and 10.3%. DRI assigns a 25% probability to future
inflation rates of 15.9%., 12.7% and 9.5%; a 15% probabil ity
is assigned to CPI rates of 15.4%, 13.5% and 15.7%; and a 10%
probability is assigned to the values of 15.4%, 12.0% and
10.5% obtaining. The means of these inflation forecasts for
1980, 1981 and 1982 are, respectively: 15.5%, 12.2% and
10.9%.
Thus, DRI's best estimate of the average rate of infla-
t ion over the next three years is 12.9%. My rather opti-
mistic forecast (14 1/2%, 10% and 9%) has a mean rate of
11. 2%; this rate may be too conservative. I expect, though,
that most economists would agree that a 11% to 12% average
rate of inflation over the next three years is probably
pretty accurate.
When an investor invests in Treasury securities, at a
minimum he must expect to earn the rate of inflation. If
not, it would be preferable to consume today rather than
invest. Thus, this range 11% to 12% represents my most con-
servative estimate of what short-term riskfree investing
will yield investors over the next three eyars.
After reporting the yields available on riskfree invest-
ing, and the terms associated with our current money markets,
it should be clear that investors minimally acceptable rates
of return over the near-term are at unprecedented high val ues.
TESTIMONY - 67
I
I
I 1
I 2
3
I 4
I 5
6
I 7
I 8
9
I 10
I 11
12
I 13
I 14
15
I 16
I 17
18
I 19
I 20
21
I 22
I 23
24
I 25
I 26
I
I
Even if the investor's impatience to consume has not increased,
both the expected level of future inflation in the United
States and the uncertainty about these future expected rates
have never been greater for perhaps the last 100 to 150 years.
It will take several years before inflation expectations are
decreased below a 10% range--even if all current economic
goals are successful. But, if util i ties are to effectively
compete in capital markets during the next two or three years,
their allowed rates of return on common equity must be sub-
stantially increased. A utility, unlike nonutilities, must
provide service at all times--not just some of the time. Given
that a utility such as Intermountain is required to provide
dependable and adequate service all of the time, it must have
the ability to attract necessary capital on fair (nonconfis-
catory) and reasonable terms in order to fulfill its obliga-
tions all of the time, regardless of the relative position of
the capital markets at any particular point in time.
I f allowed rates of return on common equity are not sub-
stantially increased, deep discounts for util i ty stocks will
ensue: low market to book ratios and high dividend yields
will obtain. But, at such low market to book ratios, util i ties
would be quite reluctant to issue new common stock and dilute
its existing stockholders' position. The predictable con-
sequence of not allowing current increases on common equity
is that utilities will increase their reliance on expensive
short-term bank loans and issue more long-term debt. The
TESTIMONY - 68
I
I
I 1
2
I 3
I 4
5
I 6
I 7
8
I 9
I 10
11
I 12
I 13
14
I 15
I 16
I
17
18
I 19
I
20
21
I 22
I
23
24
I 25
I
26
I
I
ratepayers of the future will have to bear the resulting
costly capital structure that util i ties would be forced to
create.
I think this problem is especially cri tical to Inter-
mountain. I believe that Idaho's ratepayers would be better
off if Intermountain could meet its near-term financing needs
with new equity. But this cannot be expected to occur unless
new common stock can be sold at nonconf iscatory prices. Thus,
an immed ia te and sustainahle rate of future earnings growth
is of crucial importance to Intermountain and its ratepayers.
The outcome of these proceedings will determine, in large
part, whether Intermountain can reasonably expect to issue
new common stock to meet its near term financing needs.
Q. Dr. Campbell, based on your analyses of the riskfree rate of
return, Intermountain's beta and the market risk premium,
what is Intermountain's stockholders' minimally acceptable
rate of return?
A. Intermountain's stockholders' near term minimally acceptable
rate of return is 18.5%. My best estimate of average yields
on riskfree investing is 11% to 12% over the next three year
period. Using the midpoint of this estimate, and given a
market risk premium of 8.3% and a beta of .84, the capital
asset pricing model indicates that Intermountain's stock-
holders' minimally acceptable rate of return over the near-
term is 18.5%.
Q. Dr. Campbell, you have also indicated that a more indirect
TESTIMONY - 69
TESTIMONY - 70
I
I
I 1
2
I 3
I 4
5
I 6
I 7
8
I 9
I 10
11
I 12
I 13
14
I 15
I 16
17
I 18
I 19
20
I 21
I 22
23
I 24
I 25
26
I
I
I
if the market price were lower than the maximum price inves-
tors were willing to pay. Forecasting the same expected
future cash flows at these lower market prices, the investors
would expect to receive a return greater than their minimally
acceptable rate. Realizing that this stock is a "good" buy,
the investor would attempt to purchase even more of it,
driving up the stock price and lowering the expected rate of
return. This process would continue until the expected return
just equalled the investors' minimally acceptable rate of
return. Thus, market prices would not be lower than the max-
imum price level. And, prices could not be driven to any
higher levels because, at higher price levels, and forecasting
the same expected future cash flows, the investors would
expect to receive a rate of return (on this market price)
that would be less than their minimally acceptable rate of
return. Therefore, to say that the stock market is eff icient
is to say that, on average, new investors should just expect
to receive a rate of return that is equal to their minimally
acceptable rate of return. The return expected is that mini-
mum compensation that investors currently demand for the
opportuni ties associated with riskfree investing as well as a
component that reflects the risk of the specific common stock.
I t turns out that if we knew what future cash flows the
investor expected to receive, and if we knew the current price
he paid to acquire the common stock, then we could mathemati-
cally solve for the implied minimally acceptable rate of
TESTIMONY - 71
return. This is exactly what the DCF approach does, it gives
us a mechanical way to estimate what the investor's minimally
acceptable rate of return must be. Of course, this mechanical
procedure in no way professes to describe how an investor
determines his minimally acceptable rate of return. But, it
does provide a way for an analyst to estimate what it must
have been.
Equation (4) of Schedule 8, page 2 of 2, of Exhibit 101,
sets out the formula for determining the maximum price an
investor would be willing to pay to acquire a share of common
stock in a specif ic firm. The formula, however, is not very
tractable or estimatable. In order to make the model useful,
we must make some simplifying assumptions.
If we assumed that future dividends could be expected to
increase at roughly a constant future rate, "g", then equa-
tion (4) may be rewritten as is done in equation (5) of Sche-
dule 8. Equation (5) indicates that, in order to determine
the maximum current price of a share of common stock, the
investor must estimate the future expected annual cash divi-
dend and discount this value by his minimally acceptable rate
of return, less his assessment of the future growth rate in
the firm's cash dividends. As an illustration of this formula,
if the investor's minimally acceptable rate of return was
20%, if he projected a future cash dividend of $2.40, and if
he projected a future growth rate in dividends of 8%, then he
would be willing to pay $20, at a maximum, to acquire the
TESTIMONY - 72
I
I
I 1
2
I 3
I 4
5
I 6
I 7
8
I 9
I 10
11
I 12
I 13
14
I 15
I 16
17
I 18
I 19
20
I 21
I 22
23
I 24
I 25
26
I
I
I
firm's common stock.
Working backwards, given that the current market price
in this hypothetical example is $20 per share, and given that
we could accurately estimate the investors' expectation of a
future dividend of $2.40 per share and a future dividend
growth rate of 8%, then we would have found that these facts
imply that the investor must have had a minimally acceptable
rate of return of 20%. This may be made clearer by
rearranging equation (5) to the form found in equation (6).
Equation (6) states that the implied minimally accep-
table rate of re turn can be found by add ing the inves tor's
expected future dividend yield to his present projection of
future growth. In the .prev ious hypothetical example, given a
forecast of future dividends of $2.40 per share and a present
price of $20 per share, the expected future dividend yield is
12% (2.40/20). Adding to this value the forecast of an 8%
growth rate in future dividends results in a precise estimate
of the investor's minimally acceptable rate of return--20%.
Thus, equation (6), called the DCF approach of esti-
mating the firm's stockholders' minimally acceptable rate of
return on common equity, gives us another way to check the
reI iabil i ty of the CAPM and AAA risk premium approaches.
Q. Are the DCF and the CAPM approaches inconsistent with one
another in concept?
A. No. There are no inconsistencies at all. CAPM attempts to
directly measure the investor's minimally acceptable rate of
TESTIMONY - 73
I
I
I 1
2
I 3
I 4
5
I 6
I 7
8
I 9
I 10
11
I 12
I 13
14
I 15
I 16
17
I 18
I 19
20
I 21
I 22
23
I 24
I 25
26
I
I
I
return by directly measuring the component parts that deter-
mine the minimal demands of common stockholders. DCF says
nothing about the determination of an investor's minimallly
acceptable rate of return, but through the equilbrium process
of price formulation in our capital markets, it provides us
with an indirect way to measure its currently implied value.
Q. How can equation (6) of your Sched ule 8 be used to es t ima te
Intermountain's stockholders' minimally acceptable rate of
return?
A. Due to Intermountain's rather erratic past performance, the
analysis of its past financial data would not be very useful
in determining Intermountain's stockholders' estimate of
future dividend growth.. Because of this fact, I opted to
analyze a group of gas distribution companies in order to
arrive at a range of the investors' implied minimally accep-
table rate of return.
If two firms have the same levels of risk, they should
have the same minimally acceptable rates of return. Thus,
even if the component parts of equation (6) may di f fer for
firms of like risk--the dividend yield and growth term--they
should add to the same value.
As I have previously testif ied, I believe that wi thin
varying degrees, Alagasco, Atlanta Gas Light, Bay State Gas,
Cascade Natural Gas, Gas Service Co., Indiana Gas, Laclede
Gas, Northwest Natural and Hisconsin Gas, represent utilities
that are somewhat comparable to Intermountain. I will now
TESTIMONY - 74
I
I
I 1
2
I 3
I 4
5
I 6
I 7
8
I 9
I 10
11
I 12
I 13
14
I 15
I 16
17
I 18
I 19
20
I 21
I 22
23
I 24
I 25
26
I
I
I
report two figures corresponding to each of these firms from
Value Line (April 18, 198). The first figure take the firm's
divïdend yield and adds Value Line's estimates of future
five-year growth rates in the firm's cash dividends; the
second figure is Value Line's estimate of the total return
stockholders should expect due to growth and yield. These
reported for the above firms are: 16.8% - 18%, 19.5% - 18%,
18.8% - 21%, 23.1% - 21%, 15.8% - 20%, 21.3% - 19%, 17.6% -
18 %, 17 % - 17 %, 20. 5 % - 20 %.
As is readily apparent from these data, the imp 1 ied
stockholders' minimally acceptable rate of return for these
firms is somewhere between 17 - 23 %; with a mid-point of 20 %.
Because Intermountain is among the most risky of these firms
a conservative estimate would be the mid-point value of 20%.
This result is very consistent with what the AA risk premium
analysis showed and what CAPM showed. And, it again indicates
the unprecendented high rates of return currently demanded by
stockholders, in large part due to the seriously high level
of inflation embedded in our economy.
Q. Dr. Campbell, what conclusions have you arrived at concerning
Intermountain's stockholders' minimally acceptable rate of
return?
A. The corporate AAA risk premium method suggests that minimally
acceptable rates of return for stocks of average risk, over
the next three years (1980-1982), is 19.1%.
The CAPM approach indicated that Intermountain's
TESTIMONY - 75
I
I
1
I 2
I 3
4
I 5
I 6
7
I 8
I 9
I
10
11
I 12
I
13
14
I 15
I
16
17
I 18
I
19
20
I 21
I
22
23
I 24
I 25
26
I
I
I
stockholders, or stockholders in any firm having a historic
beta of .75, must expect to receive an 18.5% rate of return
over the near term.
The DCF approach indicated that Intermountain's stock-
holders' must be conservatively demanding a mid-point value
of 20%.
The fact that all three approaches reached similar con-
clusions reinforces our confidence in the results.
Q. Do these figures represent Intermountain's fair rate of
return on common equity?
A. No. As previously explained, due to flotation costs asso-
ciated with the issuance of common stock, the stockholders'
minimally acceptable rate of return is less than the fair
ra te of return. Again,. this is exactly analogous to the fact
that, due to flotation costs on debt, a firm's coupon rate on
its debt does not represent its embedded cost of debt.
As I have previously testif ied, the fair rate of return
for Intermountain can be found by multiplying 1.06 times the
stockholders' minimally acceptable rate of return. Thus, my
best estimate of Intermountain's fair rate of return is
19.6% (18.5% x 1.06).
Q. Professor Campbell, are you advocating that the Idaho Public
Utilities Commission grant Intermountain a 19.6% rate of
return on its common equity?
A. What I am stating is that a 19.6% allowed and expected to be
earned rate of return would enable Intermountain's common
TESTIMONY - 76
I
I
I 1
2
I 3
I 4
5
I 6
I 7
8
I 9
I 10
11
I 12
I 13
14
I 15
I 16
I
I
I
I
I
I
I
I
I
stock to sell at roughly 1.06 times its book value over the
next three years. Intermountain would have the ability to
sell new common stock and improve its capital structure
wi thout confiscating wealth from its current stockholders.
Although 19.6% is my best estimate for the future three
years, due to the incredible amount of uncertainty in our
economy, a range of 17 1/2% - 22 1/2% probably encompasses
Intermountain's fair rate of return on its common equity. In
order to be ultraconservative, I suggest that a 17 1/2% rate
be used for these proceedings. As the future unfolds--
especially 1980--more precision in our forecast of the future
can be expected. At this point, with all current information,
17 1/2% is my very minimum estimate of a fair rate of return
on Intermountain's common equity.
Q. Dr. Campbell, does this conclude your direct testimony?
A. Yes, it does.
TESTIMONY - 77