1
APPENDIX B
2
THE WEIGHTED AVERAGE COST OF CAPITAL AND LEVERAGE
3
ADJUSTMENTS
4
5
1.
6
The after tax weighted average cost of capital (ATWACC) is a cornerstone of modern finance and
7
figures prominently in the testimony of Drs. Kolbe and Vilbert and the requests of Gaz Metro.
8
However, in this hearing it is used in an inappropriate way to increase “reasonable” direct estimates
9
of the fair rate of return, into recommendations that the Alberta Energy and Utilities Board in 1998
10
decided it would be derelict in exercising its statutory responsibilities to accept. In this appendix I
11
will discuss why the application of ATWACC and associated “leverage” adjustments are
12
unnecessary, how they produce such large adjustments as to violate the fair return standard and the
13
implications of National Energy Board’s 2008 TQM decision where an ATWACC approach was
14
adopted for TQM to determine tolls for 2007 and 2008.
15
It is well accepted that debt financing magnifies risk since it imposes fixed financial charges. As a
16
result several boards across Canada have allowed different common equity ratios to offset
17
differences in business risk. With the overall risk equalised, the different utilities can then be
18
allowed the same ROE.1 The regulatory approach looks at the book valued utility debt and equity
19
financing where the focus is on the debt and interest coverage ratios and the utility’s bond rating. It
20
is important to note, however, that given the similarity of allowed ROEs and capital structures the
21
market values of debt and equity should also be similar;2 meaning that investors are faced with very
22
similar overall risk. In fact this is the very basis of most estimation techniques where we use samples
23
of utilities assuming that they are relatively similar.
24
However, in this hearing Dr. Kolbe and Vilbert focus on the market value of the debt and equity as
25
their measures of financial risk and claim that because the market value of equity exceeds book
26
value for most utilities the financial risk from this lower market valued debt ratio means that the
27
equity cost estimate so derived has to be increased when applied to the book value of equity. This
1
2
Motivation for this Appendix
The NEB, BCUC, AEUB and OEB have all made capital structure changes in part to offset differences in business risk.
The market values of debt are more likely to differ across utilities given the wider range in embedded debt costs.
1
1
logic is fatally flawed and the recommendations that arise from it are totally incorrect. However,
2
before explaining this it is important to understand Dr. Kolbe and Vilbert’s approach. As Dr. Vilbert
3
states at page 9 of his testimony
4
5
The “problem” referred to is that financial leverage affects the investor’s fair rate of return and Dr.
6
Vilbert’s statement is a clear articulation of how Drs. Kolbe and Vilbert arrive at their
7
recommendations for Gaz Metro using an approach that the Alberta EUB in 1998 stated it would be
8
derelict in the exercise of it responsibilities to accept.
9
10
11
12
13
14
15
16
17
18
19
20
•
•
•
•
First Dr. Vilbert starts by stating that leverage affects the fair rate of return, which it
obviously does;
Second, Dr. Vilbert assumes that overall there is “no magic” to debt and that the ATWACC
is constant. He provides no empirical support for this statement which is patently not true;
Third he estimates the ATWACC for a sample of utilities using market value weights and
current cost estimates.
Finally with the average ATWACC the recommendations are then one of the following:
o the ATWACC is adopted with the utility free to chose their own financing;
o the constant ATWACC assumption is used to make a “leverage adjustment” to the
allowed ROE so that it can be applied to the lower book valued equity ratio;
o the constant ATWACC assumption is used to change the regulated common equity
ratio to be ‘consistent’ with a generic allowed ROE.
21
Regardless of which particular recommendation they make they all flow from the constant
22
ATWACC assumption. The logic behind the approach is tortuous, counter intuitive and plain wrong.
23
In this respect it is important not to lose sight of Dr. Vilbert’s underlying estimates of the fair rate of
24
return for a Canadian utility holding company (UHC) which first go into his WACC estimates. This
25
is the basic data” that is used in the WACC before the leverage adjustments transform the numbers
2
1
into their final recommendations. In Dr. Vilbert’s tables MJV-7 and MJV-11 he provides the
2
following estimates for Canadian UHC’s:
DCF Multi
CAPM
Canadian Utilities
7.9%
9.2
Emera
10.4
8.7
Enbridge Inc
10.2
9.9
Fortis Inc
9.1
9.5
TransCanada Corp
10.4
9.3
Average
9.60
9.32
Vilbert Average in TQM 2008
8.64
8.50
8.38
8.12
testimony
Vilbert Average in AUC 2009
testimony
3
4
The multi-stage DCF model is based on analyst growth estimates tapered to the long run growth in
5
the economy on the basis that some of the growth forecasts clearly contradict the assumptions of the
6
DCF model. The CAPM is what I refer to as a classic CAPM model using a long Canada bond yield.
7
These estimates are all biased high: the DCF estimates are based on analyst growth forecasts which
8
are known to be biased high (overly optimistic);3 while the CAPM estimates are based on higher
9
adjusted beta estimates and estimates of the market risk premium which have not reflected Canadian
10
market experience for well over fifty years. However, they are not as clearly biased as Dr. Vilbert’s
11
ECAPM results which are internally inconsistent.
12
However, there are very important implications that flow from Dr. Vilbert’s basic estimates. I have
13
averaged his five estimates in this hearing and also done the same for his estimates last year in the
14
TQM hearing (RH-1-2008) and his testimony earlier this year for the Alberta Utilities Commission.
15
What is striking is that these estimates are all much lower than the requested 12.39% allowed ROE
3
This is why I do not reference Dr. Vilbert’s estimates that simply use unadjusted analyst’s forecasts, but instead use Dr.
Vilbert’s multi-stage estimates that at least qualify the analyst’s estimates.
3
1
for Gaz Metro on its regulated 38.5% common equity ratio. This appendix’s motivation is to show
2
how estimates of 9.3-9.60% which are not too far above the Regie’s formula ROE, end up with a
3
13.39% recommendation which is outside of any reasonable range of ROE recommendations.
4
Dr. Vilbert’s estimates would be reasonable estimates of the fair return to the equity holder if these
5
UHC’s stock price were selling close to book value. In that case, all of his leverage adjustments
6
would come to naught since the market value leverage ratios in the WACC would be the same as the
7
deemed regulated common equity ratio. However, the market-to-book ratios for these UHCs are all
8
well above 1.0, which means that their market valued common equity ratios exceed their regulated
9
book valued common equity ratios. Normally a market to book ratio substantially above 1.0
10
indicates that the shareholders are very happy and the allowed ROE exceeds their fair return. This is
11
a standard implication of any book on regulatory economics, including one authored by Dr. Kolbe
12
himself.4 However, it is the genius of the methodology used by Dr. Kolbe and Vilbert in this hearing
13
that their flat ATWACC assumption converts this obvious result into exactly the opposite result:
14
they end up recommending an even higher ROE!
15
To emphasise, even though their high direct estimates of the fair rate of return are quite similar to the
16
Regie’s ROE formula return their assumption of a constant WACC and mis-use of market value
17
leverage ratios leads them to conclude that Gaz Metro should have a very significant increase in both
18
its allowed ROE and common equity ratio. I can not recall many rate hearings where a company has
19
requested both a significant increase in its common equity ratio and/or a huge increase in its ROE. A
20
request of this order of magnitude would only be justified if something cataclysmic had happened to
21
the company; yet in this case it simply comes about through using standard theoretical constructs in
22
an inappropriate way. Consequently this appendix is directed at showing how high direct ROE
23
estimates turn into recommendations that are outside of any range of reasonableness.
2 The Weighted Average Cost of Capital and ATWACC
24
25
First consider the simplest problem in finance that of a 100% equity financed firm. I will use K as
26
the investor’s required return, that is, the cost of raising equity capital, so if the firm’s earnings are a
27
perpetuity, its value V, is its forecast perpetual earnings of X divided by K or
4
L. Kolbe, J. Read and G. Hall, Estimating the rate of return for public utilities, MIT Press, 1984.
4
V=
1
X
K
(1)
2
If the earnings are expected to be $1mm in perpetuity and the typical investor requires a 10% return
3
then the firm’s value is $10mm. Note I said the investor “requires,” since this is the meaning of a fair
4
or required return as defined by Mr. Justice Lamont in Northwestern Utilities as the return required
5
by the investor to make the investment. In this sense it is also the return on other investments of
6
equivalent risk.
7
Each individual investor knows what return they require, but if we take the above equation and
8
reverse it we can infer what the typical investor requires from the current market value and the firm’s
9
expected earnings, that is,
K=
10
X
V
(2)
11
With the example numbers, $1mm in earnings valued at $10mm implies that the investor requires a
12
10% return. Since the firm has to meet this investor requirement this is the cost of raising equity
13
financing, so the 10% is also known as the firm’s cost of equity capital. In this form the equity cost
14
estimate is that made from one simply form of the DCF model.
15
The above equation illustrates the standard result in finance that market values and required returns
16
or “costs of” vary inversely. For example, we value bonds using the same formula adjusted for the
17
fact that the fixed payments do not go on in perpetuity, but instead are truncated at the bond’s
18
maturity date. However, we still get the same result: when market interest rates fall, and with them
19
investor required rates of return, then bond values go up and vice versa.
20
Further the expected earnings of the firm can be broken out into the rate of return, r, on the book
21
value of the firm’s assets A or
V=
22
rA
K
(3)
5
1
if the firm’s $1mm in earnings are the result of earning a 10% return on a $10mm book value then
2
we get the additional insight that the firm’s market to book ratio5 is simply
3
V
r
=
A K
(4)
4
In this case, since the firm is expected to earn 10% and the investor’s required rate of return is 10%,
5
the market value is equal to what the investor has contributed; that is, the firm’s market to book ratio
6
is equal to 1.0.
7
The above results are perfectly general and indicate that there are two basic ways in which a firm can
8
increase its market value: the first is by increasing its earnings through a higher rate of return r, the
9
second is through lowering its cost of capital, K. These two ways of increasing value are the heart of
10
corporate finance and reflect the value of investment and financing decisions respectively. Note also
11
that given the higher stock market value we can see that the market to book ratio now exceeds 1.0,
12
indicating that the investor has bid up the value of the shares since they are getting more than they
13
originally anticipated when the firm financed the $10mm in assets.6
14
The above implications for the market to book ratio are important for understanding how capital
15
markets work. It is a central result in finance that when investors receive more than they require,
16
market values increase, and with them the market to book ratio. Conversely, when they receive less
17
than they require; the market to book ratio drops below 1.0. This is observed every day, for example,
18
in the pricing of debt securities, where government bonds with higher coupons (interest rates) than
19
current market rates require, sell at a premium to their par value and those with smaller coupons sell
20
at a discount. In the debt markets these bonds are referred to as premium or discount bonds, but we
21
could just as easily refer to them as bonds with market to book ratios above 1.0 and below 1.0. The
22
information and meaning is exactly the same.
5
In this example the MB is simply the market value divided by the book value of total assets
The example implicitly assumes a 100% ROE regulated utility as there are more non-regulated assets it is more
difficult to look at the market to book ratio as a signal. It is then in the interests of the regulated utility to make
observing this market to book ratio as difficult as possible. I refer to this as “looking through a dirty window,”
where there is no incentive for the utility to clean the window. It may not be an accident that there are so few pure
regulated utilities left in the public capital markets.
6
6
1
I developed the examples to emphasise the importance and generality of the market to book ratio and
2
the important relationship between what the firm is earning, r, and what the stock market requires, K.
3
In the example, the firm is earning 10% and the investor requires 10%, so the stock market is telling
4
the firm “only invest in new projects that earn at least a 10% rate of return, otherwise you are
5
wasting our money.” If, for example, the firm raised a further $10mm from investors and invested in
6
a new $10mm project earning 8% in perpetuity, its new stock market value would be $18mm
$18mm =
7
$1mm + $0.8mm
0. 1
8
Its value has gone up by $8mm, but only at the cost of investing an additional $10mm, so the firm
9
has wasted $2mm of stockholder’s money. This change in value is called the net present value
10
(NPV) of the investment. Corporate finance is focussed on the firm making decisions that increase
11
shareholder value and not waste it. The key message is that the firm should only invest in projects
12
earning at least the firm’s cost of capital, which in the example is 10%.
13
Now suppose market interest rates drop and with them the investor’ required rate of return. This has
14
been one of the central features of capital markets over the last twenty years. Long-term interest rates
15
have been on a long-term downward trend since peaking at over 18.0% in September 1981. Suppose
16
in our example the firm’s cost of capital drops from 10% to 7%. In this case, the firm’s market value
17
would increase to $14.3mm and the market to book ratio increases to 1.43X. Again the market to
18
book ratio indicates that the firm is now earning 10%, when investors only require 7%. However, the
19
original investors have earned an unexpected 43% capital gain, since they only expected to earn a
20
10% rate of return through the $1mm in earnings on their $10mm investment. If this were a utility
21
the stockholders would have earned an excess return and we can note this by looking at the market to
22
book ratio: any market to book ratio significantly above 1.0 indicates that investors have in the past
23
earned a return that is above a fair and reasonable rate of return.7
24
Further, the 8% investment that previously destroyed $2mm in value now increases it, since 8%
25
exceeds the new 7% cost of capital. The new value for the firm with the investment is
7
If the firm is sold for $14.3 mm, new investors only earn their 7% required return, since their return is based on
the $14.3mm investment cost, that is., $4.3mm of “goodwill” plus the $10mm original cost.
7
$25.7 mm =
1
$1mm + $0.8mm
0.07
2
This $25.7mm is the $14.3mm market value of the original investment, plus the $10mm cost of the
3
new investment plus an additional $1.43mm NPV from the new investment.
4
The example indicates that the firm has to constantly monitor its cost of capital: a project that it
5
would not accept when its cost of capital was 10%, it will accept when it drops to 7%. Moreover,
6
this investment yardstick is NOT a rate of return earned by other firms, that is, other firms “r,” but
7
the market cost of capital, K. The rate of return, r, earned by other firms is irrelevant in corporate
8
finance, since the firm has to satisfy its stockholders, not other firms.8
9
These basic principles are exactly the same when the firm finances with debt. Suppose the firm
10
continues to have $10mm in assets earning 10%, but now there is $5mm in debt at a cost of 5% and
11
$5mm in equity at a cost of 15%.9 The firm’s weighted average cost of capital (WACC) is the value
12
weighted average of these two sources of capital. In this case 50% debt at a cost of 5% and 50%
13
equity at a cost of 15% gives a WACC of 10%. However now that the firm has debt and equity we
14
have to distinguish between its equity market value and its total enterprise value. First, define the
15
rate of return on the firm’s assets as its return on investment (ROI), which is the return prior to
16
meeting interest payments divided by total assets. In the example, the ROI is 10% so the firm
17
continues to earn $1 million before making any interest payments, these earnings are referred to as
18
EBIT: earnings before interest and taxes. Now we subtract the interest of $0.25mm (5% times
19
$5mm) from EBIT to get net income of $0.75mm.10 If this $0.75mm in net income is discounted at
20
15%, then we get the equity market value of $5mm. That is,
$5mm =
21
$1mm − $0.25mm
0.15
or algebraically
22
8
This is the core reason why comparable earnings estimates, which simply try to estimate r, are of no value in
public utility regulation.
9
15% is used simply for ease of calculation.
10
There are no taxes yet.
8
$5mm =
1
ROI * A − K b * B
Ke
2
where A is the total book value of assets, B is the amount of debt financing ($5mm) and I have
3
subscripted the two costs, b for debt (bonds) and e for equity.
4
In this example the market value of the equity is $5mm so that the total enterprise value (V), or
5
overall market value of the firm (debt plus equity), is the $5mm equity value plus the $5mm of debt
6
or $10mm. This calculation is an example of the flows to equity (FTE) method of valuation, where
7
the flows to the equity holder are discounted at the cost of equity capital to directly determine the
8
value of the equity. However, to do this calculation we need the value of the debt financing and most
9
corporate investment decisions are separated from these financing decisions. Consequently, it is
10
conventional to rearrange this equation to get the WACC. First multiply through by the cost of
11
equity,
E * K e = ROI * A − K b * B
12
13
where I have substituted E for the equity market value. Second, group the equity and debt costs and
14
factor for the overall market value to get,
V (K e
15
16
dividing through we get
V =
17
E
B
+ K b ) = ROI * A
V
V
ROI * A
E
B
Ke + Kb
V
V
(5)
18
where the total enterprise or market value is equal to the EBIT discounted at the weighted average
19
cost of capital (WACC). In our example the $1mm in EBIT discounted at the WACC of 10% gives
20
the total market value of $10mm. The equity value is then this $10mm value minus the $5mm in
21
debt or the same $5mm as calculated using the flows to equity approach.
9
1
The WACC simply recognises the different sources of finance available to the firm and averages
2
them to get an overall cost of capital. In this sense the cost of capital is a blended cost of financing to
3
the firm, but once this is done all the previous results hold just as before. For example, if the ROI
4
exceeds the WACC then the market value increases and the market to book ratio is again above 1.0.
5
For example if the ROI increases to 11% and the WACC stays at 10% then the total enterprise value
6
increases to $11mmm. The equity value is then $11mm minus the $5mm in debt, so the equity value
7
increases to $6mm and the equity market to book ratio increases to 1.2X. Similarly, if the cost of
8
equity declines to 13% from 15%, then all else constant, the WACC drops to 9% and the market
9
value increases to $11.11mm. Again with $5mm in debt the equity value increases to $6.11mm for a
10
market to book of 1.22X. Finally, if the WACC drops to 9%, this becomes the hurdle rate for new
11
investments; that is, the ROI on new investments has to be greater than 9%.
12
Adding corporate income taxes does not materially change any of the basic results. All that happens
13
is that the EBIT becomes taxable, while the interest expense is tax deductible. In this case it is this
14
after tax net income that is discounted by the equity cost, that is,
Equity =
15
( ROI * A − K b * B)(1 − T )
Ke
16
where T is the corporate tax rate. Rearranging, as before, means that the after tax ROI has to exceed
17
the after tax WACC or
V =
18
ROI (1 − T ) * A
E
B
K e + K b (1 − T )
V
V
(6)
19
The only difference is that since interest is tax deductible, whereas equity costs are not, the after tax
20
ROI has to exceed an after tax WACC or ATWACC.
21
It is fundamental to corporate finance that the ATWACC uses market values. This means for
22
example, that the debt and equity ratios use market or target values for debt and equity divided by
23
the total enterprise value and not book values. The ATWACC then gives the current yardstick that the
24
firm has to beat in order to create shareholder value, that is, to increase the firm’s market value. Only
10
1
by calculating the ATWACC in this way can the firm be sure that it is accepting projects that enhance
2
shareholder value, rather than destroying it.
3
3. ATWACC, WACC and the Regulation of Utilities
4
It is important to recognize that the ATWACC is critical for the concept of shareholder value
5
maximization: if the firm is not expected to earn its ATWACC then its market value will fall. The
6
ATWACC is thus a critical concept to understand how firm’s can make decisions that enhance their
7
market values. In contrast regulators are not concerned with maximising or enhancing shareholder
8
value; their mandate is to set “fair and reasonable” rates. This frequently puts them at odds with
9
maximising shareholder value, since regulation should never be designed to rubber stamp market
10
values. What this means is that the regulator sets rates and through them the firm’s EBIT and ROI,
11
whereas for non-regulated firms the ROI is determined in the marketplace.
12
To continue with the previous (no tax purely for simplicity) example, where the WACC is 10% and
13
the equity cost 15%, suppose the regulator institutes some risk reduction techniques such as the use
14
of a forward, instead of an historic test year,11 or the use of deferral accounts. As a result, the equity
15
cost drops to 11% due to the reduction in risk. Now, everything held constant, the debt is still valued
16
at $5mm and the rate base (total assets) $10mm; so the only thing that has changed is the equity cost.
17
In this case, the equity value can be determined from the flows to equity formula as
$6.818mm =
18
$1mm − $0.25mm
0.11
19
Note that the equity holders recognise the reduction in risk so they bid up the stock market value
20
from $5mm to $6.818mm due to the lower required rate of return. In this case the existing
21
shareholders get an additional capital gain of 36.4% over and above their fair return.12 If the firm is
22
now in a rate hearing to adjust its ROE, a tip off to the regulator is that the market to book ratio has
23
gone from 1.0 to 1.364X (6.818/5), so intuitively by lowering the firm’s risk and seeing the market
24
value increase the regulator knows that the allowed ROE has to be cut. The obvious thing for the
25
regulator to do is simply get expert opinion to estimate the equity cost, and if this is unbiased, notice
11
Forward test years remove any inflationary bias involved in the use of an historic test year adjusted for specific
identifiable changes. With the decline in inflation most of the need for forward test years is removed.
12
As perpetuities they get their fair return as the earnings are paid out as a dividend.
11
1
and cut the allowed ROE to 11%. The equity value will then return to $5mm and the stockholders
2
will continue to earn a fair return on their $5mm investment.
3
The question is then what does estimating the WACC add to this process? Assuming there is no bias
4
to estimating the equity cost at 11.0% the new WACC is
WACC = 0.11 *
5
6.818
5
+ .05 *
11.818
11.818
6
or 8.46%.13 The most important thing to note is that the WACC uses market value weights and since
7
the equity market value has gone up to $6.818mm, the WACC uses an equity ratio of 57.7% and a
8
debt ratio of 42.3%, rather than the assumed regulated weights of 50% for both.
9
The reason for the use of market value weights is that the WACC is the minimum rate of return the
10
firm has to earn to maintain its market value, which has increased from $10mm to $11.818mm.
11
Theoretically, it makes no difference whether this $11.818mm is the result of actually raising
12
$11.818mm, or whether it is simply the current market value of the original $10mm investment as it
13
is in this case. The point is simply that using WACC as a cut off rate reflects what the firm has to
14
earn to sustain this current market value, that is, “keep what it has got.” In particular, the new WACC
15
of 8.46% is applied to the market value of $11.818mm.
16
In contrast, the regulator should not be interested in letting the investors “keep what they have got”
17
instead, they have to ensure that rates are fair and reasonable. In the example it is clear that the
18
allowed ROE has to be cut and implicitly that the equity market value has to fall. Moreover, the
19
regulator has to determine a fair return on the book value of the investment, that is, the rate base, not
20
the firm’s market value. In this sense, there is a fundamental contradiction to applying the
21
conventional WACC or ATWACC to rate of return regulated firms.
22
However, suppose the Board tries to apply WACC. First, note that this exercise is much more prone
23
to error, and as a result subjective, than just estimating the fair return directly. This is because as
24
well as estimating the equity cost, you have to estimate the market, not the embedded cost, of debt
13
Note that discounting the $1mm in pre interest earnings by 8.46% gives the total enterprise value of
$11.818mm. This is the new cut off rate for evaluating the firm’s investments.
12
1
the financing weights and the appropriate tax rate. All of these components are subject to error, since
2
many issues of debt are not traded and as a result it is difficult to estimate either their cost or their
3
market value. However, assuming all these additional estimation problems can be solved, suppose
4
the correct 8.46% WACC is estimated and awarded the regulated firm as NGTL requests, what
5
happens next?
6
If the regulator accepts this WACC and applies it to the book value rate base, the equity value drops
7
to $5.42mm or
V =
8
.0846 * $10 − .25
.11
9
Although the ROI is reduced from 10% to 8.46%, it is not reduced to the correct ROI of 8.0%,14 so
10
the equity market value is still $0.42mm higher than it should be. The reason for this is that using
11
market value weights in the WACC puts greater emphasis on the higher equity cost than the debt
12
cost. For non-regulated firms this is correct, since the objective is to maintain these higher market
13
values. However, it is totally incorrect for a regulator who is tasked with awarding a fair return
14
regardless of what happens to the stock price. By estimating and applying a market based WACC to
15
a book value rate base the effect of the higher allowed ROE is perpetuated through its impact on
16
the higher equity market value.
17
Over time, if nothing else changes, the excess value will gradually be removed. For example, as the
18
market value falls to $5.42mm the new WACC becomes
WACC = 0.11 *
19
5.42
5
+ .05 *
10.42
10.42
20
or 8.12%. Again if everything remains constant, in the next rate hearing the regulator would cut the
21
allowed ROI to this level and the equity market value would fall again until after successive rounds
22
it ends up at the $5mm fair value for the equity. Note that the regulated firm, as well as others with a
23
vested interest in the firm as an investment, may complain about the regulator being tough by
24
repeatedly cutting the allowed ROE, but it is not being tough at all. The ROE is still above the fair
14
The correct regulated WACC is the average of the debt and equity costs using regulated book value weights, in
this case 50%.
13
1
ROE. However, by using market value weights in the WACC and by shifting the focus from the ROE
2
to the WACC, this adjustment process is drawn out to the stockholders’ benefit. Further it gives the
3
regulated firm an opportunity to bring up other arguments that may delay even this adjustment.
4
Consequently the adoption of WACC based regulation delays the adjustment process to the
5
stockholder’s benefit.
6
The basic insight from this discussion is that by using market values in WACC, the resulting cost of
7
capital is higher than a fair return, since the higher equity cost is given a greater weight. Further if
8
the firm is a pure ROE regulated utility it tends to “rubberstamp” the use of market values and is
9
contrary to “fair and reasonable” regulation. This is because the market to book ratio is the basic
10
signal as to whether or not investors are being treated fairly. When we add in flotation costs and a
11
desire to allow the regulated firm to access markets at all time, most would target a market to book
12
ratio marginally above 1.0, say 1.10. However, apart from this minor deviation from book values,
13
the essential point is that the correct financing weights for a regulated firm should be the regulated
14
capital structure weights, not the market value weights. To incorporate into the regulatory process a
15
regulated firm’s market value is to rubberstamp investor expectations, however unrealistic, and delay
16
the adjustment to a fair and reasonable value for the allowed ROE.
17
The Alberta EUB directly addressed the use of ATWACC on a number of occasions. For example,
18
in connection with comparable earnings testimony the EUB stated (Generic Cost of Capital Decision
19
U-200452, page 24)
20
21
“The Board considers that the application of a market required return (i.e. required earnings on
market value) to a book value rate base is appropriate in the context of regulated utilities.”
22
That is, you estimate a market opportunity cost, such as that from the CAPM, and apply it to book
23
values, not market values as is the assumption in WACC.
24
In explicitly considering the usefulness of ATWACC the EUB stated (Decision U-99099, page 300)
25
26
27
28
“The Board observes that the intrinsic long-run value of a pure play regulated entity is best
represented by book value. In other words, the present worth of future regulated earnings,
discounted at the allowed return, is by definition equal to book value assuming achieved
regulated earnings on average equal allowed regulated earnings. Accordingly, the Board
14
1
2
considers that book capitalization represents the best indicator of the long-run market
capitalization for a pure play regulated firm.”
3
It is difficult to see how a regulator could say anything other than what the EUB stated above, since
4
to accept a market to book ratio much above 1.0 is in effect to rubberstamp unrealistic investor
5
expectations or to admit that allowed ROEs are too high. The EUB further recognised this when it
6
went on to say (U99099, page 303)
7
8
9
“The Board would be derelict in its statutory responsibilities to recognize market capitalization
ratios that are derived from a market value capitalization that deviates from the intrinsic long-run
value of the regulated firm.”
10
This is the clearest possible statement by a regulator of the fundamental grounds for rejecting
11
ATWACC and its emphasis on market values.
12
Further the EUB went on to say
13
14
15
16
17
18
19
20
“In essence, a regulated company’s earnings are driven by the portion of the original cost rate
base deemed to be financed by common equity. This fact results in a fundamental disconnect to
the theory that market capitalization ratios, which have deviated significantly from book
capitalization ratios, reflect the appropriate financial risk necessary to determine a fair composite
return to be applied to the original cost rate base of a pure play regulated utility. This is because
the earnings of a pure play regulated utility are governed by and driven by the regulated return
allowed on book equity. In other words, it is the book equity that reflects the appropriate
financial risk necessary to determine a fair composite return for a pure play regulated utility.”
21
This means that the correct financial risk measure for regulated utilities is the book debt equity ratio
22
and not that determined using market values. It is also the approach pioneered by the National
23
Energy Board, where financial risk adjustments using the deemed common equity ratio (based on
24
book values) are used to adjust for differences in business risk.
25
The EUB went on to calculate an ATWACC using regulated book value capital structure weights
26
and the embedded debt costs. In this case (Decision U-99099, page 303)
27
28
29
An ATWACCBV would be suitable for a regulated utility whose profit, by legislation, is limited
to a fair return on the book value (i.e. original cost) of its assets. The Board notes that an
ATWACCBV is consistent with the logic of the traditional method of determining fair return.
15
1
In our example, the ATWACCBV is the 5% debt and 11% equity cost weighted with the 50%
2
regulated capital structure weights. In this case the ATWACCBV is 8.0% and awarding this 8% cost
3
of capital means that the value of the equity is
V =
4
.08 * $10 − .25
.11
5
or $5mm. This is the exact same result that would arise if the regulator simply awarded the 11%
6
ROE directly.
7
The EUB ATWACCBV correctly recognised that apart from any estimation error attached to the
8
equity cost, the WACC approach is totally inconsistent with allowing a fair return to a regulated firm.
9
The only approach consistent with allowing “fair and reasonable” rates is to estimate a sample of
10
comparable firm’s ATWACC using book value weights and embedded debt costs. In this case the
11
exercise comes down to the normal problem of whether or not the estimated equity cost is accurate
12
or not.
13
4. The Need for Leverage Adjustments
14
The final step in the process used by Dr. Kolbe and Dr. Vilbert is to adjust for differences in the
15
financial leverage between the calculated WACC estimates and the firm in question. That is, given
16
the use of market value weights in the calculated WACC, which in my example were 57.7%
17
common equity and 42.3% debt, are leverage adjustments needed to apply the estimates to the
18
regulated book equity, which in our example was 50%?
19
Note that in this example the higher equity value came about by construction because of a drop in
20
risk and the allowed ROE should be cut. That is, the firm did not substitute equity for debt and
21
reduce its financial risk. Moreover, without any change in interest payments there is no change in
22
financial risk. The fact is that there is a big difference between the impact of substituting equity for
23
debt in a regulated firm’s capital structure and an increase in equity market value caused by a decline
24
in the fair return.
25
Further in the example the equity is obviously riskier at $6.818mm with a common equity ratio of
26
57%, after it has increased in value, than it was before at the regulated 50%. This is because it is
16
1
highly unlikely that the regulator will cut the allowed ROE when the stock is trading at book value.
2
In contrast, the higher the market to book ratio the more likely the regulator will cut the allowed
3
ROE and thus the riskier the stock. In contrast, Drs Kobe and Vilbert would have us believe that the
4
equity in a regulated firm is less risky when it is trading at a market to book well above 1.0, since the
5
debt ratio is lower. This simply does not make any sense.
6
It is important to remember that the financial leverage risk premium stems from the imposition of
7
fixed interest charges. That is, prior to receiving their equity return the firm has to pay these interest
8
charges. This risk does not change as the market value of the firm changes; it only changes when
9
book values change. For example, if Gaz Metro moved to a regulated (book) 50-50 debt tol equity
10
ratio there would be less interest expense. Consequently, the financial risk, both to the bond-holder
11
and the stock-holder, would decline and a leverage adjustment would indicate this. As Standard and
12
Poors have stated,
13
14
15
16
“Similarly ratios using market value of a company’s equity in calculations of leverage are given
limited weight as analytical tools. The stock market emphasises growth prospects and has a short
time horizon; it is influenced by changes in alternative investment opportunities and can be very
volatile. A company’s ability to service its debt is not affected directly by such factors.”
17
That is, S&P is basically saying book value leverage is important for assessing the default or credit
18
risk, whereas market values don’t count. If it is book values and interest payments that affect credit
19
risk and the cost of debt, then this is the risk that also affects equity investors not market values.
20
Following the AEUB’s decision that to accept market value weights would be a “dereliction” of
21
duty, the obvious implication is that the weights in the sample WACC should also be book value
22
weights. In my example this means that the regulated book value of 50%, rather than the market
23
value debt ratio of 42.3% is what matters. Hence in comparing this 50% regulated debt ratio with the
24
firm in hand that also has 50% debt ratio means that no adjustment is necessary. Making an
25
adjustment based on market values is then inappropriate for a regulated firm. As the AEUB again
26
noted (Decision U99099, page 301)
27
28
29
30
“the Board considers that beta and the cost of equity do not change to the extent necessary for an
ATWACC, determined from market capitalization weights, to remain constant when applied to
the book capitalization for a pure play regulated utility. The increase required to the cost of
equity to achieve a constant ATWACC would be excessive and violate the fair return standard.”
17
1
It is my judgment that the only time a leverage adjustment is needed is either when the overall risk
2
differs between the utility and the sample of regulated firms or when its business risk changes and
3
the Board wants to offset this change so it can continue to award a formula allowed ROE. In these
4
cases, the Board is best advised to base its decision on business risk and financial market access as it
5
has done in the past.
6
5. The Size of Leverage Adjustments
7
It is well accepted that financial risk magnifies business risk. The basic financial leverage equation
8
indicates that the accounting return to the stockholder is determined as follows
ROE = ROI + ( ROI − Rd ) D S
9
(7)
10
where these are all book values, that is, D and S are the book values of debt and equity and Rd is the
11
embedded cost of debt. The equation simply comes from manipulating the firm’s financial
12
statements. It means, for example, that with a fixed stock of assets, as revenues and the ROI change,
13
the greater the amount of debt the greater is the variation in the accounting return to the stock
14
holders. However, the above equation says absolutely nothing about how the stock market reacts to
15
this financial risk, that is, how market values change, or how the cost of equity changes as the firm
16
uses debt.
17
To understand how the investor’s required rate of return or equity cost varies with the use of debt we
18
need a valuation model. The first valuation attempt was by Franco Modigliani and Merton Miller
19
(M&M) who in 1958 developed an arbitrage model to show that the total enterprise value was
20
independent of the use of debt. This was their famous “no magic in debt argument.” If individuals
21
can borrow on the same terms as the firm, then investors will not pay a premium for firms that use
22
debt. In this case corporate leverage is identical to corporate leverage and the firm adds no value
23
by using debt. Consequently, they derived the following formula
24
K e = K 0 + (K 0 − K b ) B E
18
(8)
1
where the K’s indicate the cost of equity and debt, that is, fair or required returns and not accounting
2
returns, and B and E then represent the market values of debt and equity respectively. The subscript
3
0 then indicates what the equity cost would be if the firm had no debt outstanding, this is often
4
referred to as the unlevered equity or the asset cost.
5
Note two things about this equation. First, apart from redefining returns and debt ratios, in form it is
6
the same as the financial leverage equation. This is because in the accounting model total assets are
7
fixed, whereas in this valuation model M&M proved that the value of the firm was fixed and
8
independent of leverage under their assumptions. As a result, changes in the book and market debt
9
ratios have the same impact. Second M&M “proved” that as the market value was constant the
10
weighted average cost of capital was also constant, which in this case means that it is equal to the
11
unlevered equity cost. However, note that I italicised proved, since this was a mathematical proof
12
that followed from their assumptions, not a description of reality.
13
In the M&M equation changes in the market valued debt equity ratio (B/E) are multiplied by the
14
spread between the WACC and the cost of debt. It is this coefficient that determines how much
15
changes in the debt equity ratio affect the equity cost since it is this coefficient that determines the
16
risk. This is the important point: people who believe that changes in the debt equity ratio have a big
17
impact on the equity cost believe that the coefficient on the market valued debt equity ratio is high
18
and vice versa.
19
However, the overall market value in the M&M model is only fixed by their assumptions.
20
Remember from equation (6)
V =
21
After − tax operating income
WACC
(9)
22
for a perpetuity, the total enterprise or firm value is after tax operating income divided by the after
23
tax WACC. Given that M&M were discussing capital structure not operating changes, they assumed
24
that the after tax operating income, the numerator above, was constant. What M&M then “proved”
25
was that with firm value constant the WACC must also be constant. In this case, given that the
26
WACC is a weighted average of the debt and equity costs, the equity cost has to increase with more
27
debt to offset the impact of more “cheaper” debt. This is what equation (8) indicates.
19
1
However, if the market value increases with more debt then from equation (9) the cost of capital
2
must decrease and vice versa. In this case, the equity cost may then increase or possibly even
3
decrease with the use of debt, it all depends on the valuation model and the assumptions that are
4
made. The critical question is how the use of debt affects the overall firm value; the impact on the
5
WACC and the equity cost then follow directly.
6
M&M’s “no magic in debt” result was controversial in 1958 and remains so today. This is because
7
of the assumptions required to “prove” their result. The most important are that:
8•
9•
10•
11•
12
13•
14•
15•
16•
there are no taxes of any kind;
there are no transactions costs;
there are no information asymmetries between borrowers and lenders;
everyone can borrow on the same terms, that is., if the company can issue 25 year bonds or
access the swap market, then so too can other individuals;
all firms are perpetuities that pay out 100% dividends;
there are no bankruptcy or financial distress costs;
there are two or more identical firms with different levels of debt that can be arbitraged.
the managers operate the firm to maximize enterprise value and not their own personal goals.
17
A result of all these assumptions is that personal and corporate leverage are identical. All of these
18
assumptions have been disputed to a greater or lesser extent and if any of them are incorrect then the
19
total value of the firm is affected by the use of debt and so too will its cost of capital (WACC).
20
M&M’s result is a classic in corporate finance and they won the Noble prize in economics for it.
21
However, its great strength lies not in its result, which few accepted then or now, but the fact they
22
focused corporate finance on the implications of their assumptions. For example, in 1963 they
23
recognised that they made a mistake in their treatment of corporate income taxes and corrected their
24
original paper. They then showed that, all else constant, the value of the firm increases due to the tax
25
shield generated by the tax deductibility of interest payments. The reason is simply that what we
26
term value is the private value and by reducing corporate income taxes the private value of the firm
27
increases at the expense of the government. Hence from equation (9), if the private market value
28
increases due to the tax shield value the WACC of necessity must decline.
29
In fact in the M&M (1963) model the WACC declines continuously, and the equity cost changes as
30
follows,
20
K e = K 0 + (1 − T )( K 0 − K b ) B E
1
(10)
2
There is still a financial leverage risk premium but it is now smaller, since the use of debt also
3
generates a valuable tax shield. Note that in equation (10) the debt equity ratio is now multiplied by
4
(1-T), since part of the interest payments are paid for by the government through lower tax
5
payments. Assuming a 40% corporate tax rate, people who believe in M&M (1963) would estimate a
6
leverage impact only 60% the size of those who believe in M&M (1958). A different model
7
produces a different leverage adjustment!
8
Since 1963 all the other assumptions of M&M have been relaxed and every time an assumption has
9
been relaxed there is another leverage equation similar to equations (8) and (10) and another estimate
10
of the leverage adjustment. However, two main theories of capital structure have emerged: the static
11
trade off (STO) model and the pecking order hypothesis (POH). The STO is a static model that
12
assumes that firms trade off the tax advantages of using debt against the loss of financial flexibility
13
that arises due to excessive leverage. It is this model that develops the familiar “U” shaped WACC
14
function below as the firm increases its debt ratio.
The U shaped WACC
15
WACC
16
17
18
19
20
Significant tax
advantages
Loss of financial
flexibility severe
risk of distress
21
21
1
Initially the WACC declines due to the tax advantages of debt. In the M&M (1963) model, for
2
example each dollar of debt increases the firm’s market value by the value of the corporate tax rate;15
3
the WACC then starts to increase as the loss of financial flexibility sets in. Obviously there has to be
4
some offset to the tax deductibility of interest, otherwise all firms would try to finance with 100%
5
debt. The offset comes as the debt becomes riskier and has to be sold on higher and higher yields and
6
the firm loses its financial flexibility.
7
In contrast, the pecking order hypothesis (POH), developed in 1963 by Gordon Donaldson at
8
Harvard, is a dynamic model of financing based on the fact that firms are controlled by managers. In
9
this case, firms raise capital by issuing securities that impose the least restrictions on management.
10
Consequently, firms primarily rely on internal funds and only after these are exhausted do they go
11
outside for capital, where they initially rely on bank debt and bonds, rather than new equity.
12
I have reviewed these basic ideas on capital structure since the flat ATWACC approach of Drs,
13
Kolbe and Vilbert is essentially the 1958 M&M model extended to include corporate and personal
14
taxes by Miller (1977). This is a very important model and for the last 31 years I have invariably
15
taught corporate financing to second year MBAs with the first five weeks devoted almost
16
exclusively to these ideas,16 as well as to the implication that if this model holds there is no value to
17
the activities of investment bankers and they should all study marketing! I then spend the balance of
18
my course explaining how companies add value by adopting different financing decisions. The fact
19
is that financial theory has come a long way since 1958 and is now better harmonised with practise:
20
no one believes the flat WACC model fits reality; it is simply a good starting point to discuss how
21
investment bankers can create value for firms.17
22
However, a flat ATWACC does have the advantage that it gives just about the largest possible
23
leverage effect, that is, the coefficient on the market valued debt equity ratio in the equity cost
24
equation 8 is as large as possible. I showed earlier that the M&M 1958 flat WACC model gives a
15
This simple model has been dubbed adjusted present value (APV) by Professor Myers. In Principles of
Corporate Finance (2nd Canadian edition, 1991 pages 490-493 they work an example and the base case NPV of
$170,000 is then increased by $592,000 by the tax advantages to debt. In this case, Professor Myers, who Dr.
Kolbe references throughout his testimony, clearly believes in the tax advantages of debt.
16
This is MGT2300. A course outline can be downloaded from my web page at
http://www.rotman.utoronto.ca/~booth
17
It would be interesting to ask why investment bankers are so well paid if corporate financing decisions as
represented by a flat ATWACC have no value.
22
1
bigger equity cost adjustment (equation (8)) than if the WACC declines with leverage in the
2
conventional way (equation 10). However, Dr. Kolbe goes further by assuming a flat ATWACC in
3
the presence of corporate taxes, which gives an even bigger coefficient on the market valued debt
4
equity ratio. Even though there is nothing in his discussion that indicates that he has included the
5
non-tax offset to the tax advantages such as financial distress and bankruptcy.
6
To illustrate, Drs. Kolbe and Vilbert get their leverage adjustment by assuming a flat, that is,
7
constant WACC. Dr. Vilbert first calculates the WACC using market value weights from his
8
sample:18
Ke
9
E
B
+ K b (1 − T ) = WACC = K A
V
V
(11)
10
Drs. Kolbe and Vilbert then assume that the WACC (KA) is constant and then either alter the equity
11
ratio to get a new equity cost or the equity cost to get a new equity ratio to get the same WACC. In
12
terms of the equity cost, implicitly Dr. Kolbe is rearranging this WACC equation to solve for the
13
equity cost (Ke) at any leverage ratio,
K e = K A + ( K A − (1 − T ) K b ) B E
14
(12)
15
Since the WACC is assumed constant, it should equal K0, so the main difference is that it is this cost
16
minus the after tax cost of debt that determines the coefficient on the debt–equity ratio. This is the
17
coefficient that determines how the fair return (the equity cost) should vary as the debt equity ratio
18
changes or alternatively how the debt equity ratio has to change given a fixed fair return.
19
With a constant WACC this coefficient on the debt equity ratio which is called the leverage
20
coefficient, is larger than either the M&M (1958) no tax case or the M&M (1963) tax case as a
21
simple comparison with equations (8) and (10) indicates. In fact, as far as I am aware it is the largest
22
coefficient possible, as I have not seen any other equity cost or fair return equation with a larger
23
coefficient.
18
As explained earlier the use of market values is not appropriate for regulated firms, either directly or indirectly
through WACC estimates from samples of regulated firms.
23
1
The reason for the very large leverage adjustment in equation (12) is that the model is internally
2
inconsistent. Equation (12) and the flat WACC assumes the tax deductibility of interest, which causes
3
the WACC to fall, but there is no explicit account of the offsetting costs that negate this to keep the
4
WACC constant. For example, if the WACC is constant it could be that as the market valued debt
5
equity ratio increases the debt cost also increases due to the higher risk of insolvency and the costs of
6
financial distress and bankruptcy. In this case, what is keeping the WACC constant is an increasing
7
Kb as creditors protect themselves from the insolvency risk attached to highly debt financed firms.
8
Moreover, it is obvious from equation (12) that if the debt cost, Kb, increases with the debt equity
9
ratio then the equity cost does not increase as quickly, which is what Solomon showed in the Journal
10
of Finance in 1963.19 The intuition is simply that “debt” in highly debt financed firms has some of
11
the same characteristics as equity.20
12
To show these principles return to the previous example, where the equity cost was assumed to
13
decrease from 15% to 11% due to a reduction in risk. As a result the equity market value increases
14
from $5mm to $6.818mm and the market valued debt ratio decreases from 50% to 42.3%. Drs.
15
Kolbe and Vilbert would then say that any direct equity cost estimate at this lower market debt ratio
16
has to be increased when applied to the higher book debt equity ratio. So the question is: what is the
17
coefficient that we multiply the increased debt equity ratio by to get the higher fair return due to the
18
“greater” book leverage? Although there is NO need for a leverage adjustment as the equity cost is
19
accurately estimated at 11.0%, how could one be made?
20
One way is to estimate an unlevered equity cost from equation (8) by inserting the debt cost of 5%,
21
the debt equity ratio of .423/.577 or 0.7333, and the equity cost of 11%. In this case, the unlevered
22
equity cost is 8.46% and the use of debt financing has increased the equity cost from the debt free
23
8.46% to the observed 11.0%, so 2.54% is the financial leverage risk premium. The coefficient on
24
the market valued debt equity ratio in this example is 3.46% (8.46-5.0). The relevered equity cost at
25
the 50:50 debt equity ratio would then be 11.92%. So someone believing in M&M (1958) would
26
calculate the increased debt-equity ratio of 0.266 (1.0-0.733) and multiply by this leverage
19
The only reason for the cost of debt to increase is the risk of financial distress or bankruptcy, which M&M
ignored in their 1958 paper. Therefore, Solomon’s result is inconsistent with the M&M assumptions. However, it
is consistent with a model of bankruptcy and financial distress.
20
It could be that Dr. Kolbe has the Miller (1977) equilibrium in mind, but this is a US model that relies on
arbitrage in the non-taxed municipal bond market. There is no equivalent market in Canada.
24
1
coefficient of 3.46%. Further if they believed that the equity cost estimated from a sample of firms
2
with lower market valued debt ratios underestimated the financial risk at the regulated firm’s debt
3
ratio, they would increase the 11.0% by 92 basis points.
4
If instead the M&M (1963) with taxes equation (10) is used with a 50% tax rate, the unlevered
5
equity cost is higher at 9.39% and the financial leverage risk premium is only 1.61%, since part of
6
the interest costs are paid for through a reduction in taxes. As a result, the financial leverage risk
7
premium is only half what it is with the flat WACC M&M 1958 model. In this case the coefficient on
8
the market valued debt equity ratio is 2.20 ((9.39-5.0)*.5). Relevering to the 50% debt ratio increases
9
the equity cost to 11.59 or 33 basis points less than by using the flat WACC M&M 1958 model.
10
Believing in M&M (1963) gives a smaller bump to the ROE estimate by making leverage
11
adjustments
12
Believing in a constant WACC gives a WACC and unlevered equity cost of a constant 7.4%.21 Hence
13
the market valued debt equity ratio is multiplied by (7.4-2.5) or approximately 5.0%. This is higher
14
than either M&M (1958) no tax or M&M (1963) with tax and gives the highest possible leverage
15
adjustment. This is because the debt cost is after tax and there are no explicit offsetting costs in the
16
model, yet the WACC is somehow held constant. Using this model the leverage adjustment would
17
not be 59 or 92 basis points but 131 basis points to move the equity cost at the regulated debt ratio to
18
12.31%.
19
Let me make the importance of this example clear. The chain of events is that the risk of the utility
20
has declined causing its equity cost to drop from 15% to 11%. The obvious thing that the regulator
21
should do is cut the allowed ROE from 15% to 11%. This is also what would happen if the regulator
22
used the AEUB’s ATWACCBV approach and recognised that it would be “derelict” in exercising its
23
statutory responsibilities by using market values. However, using the “(AT)WACC approach”
24
avoids this ROE drop in two ways. The first is to go directly to the WACC with market values, which
25
seals in the higher equity ratio and delays the drop in the allowed ROE. However, if this fails the
26
second step is to argue for a leverage adjustment. Then the assumption of a flat ATWACC generates
27
just about the biggest coefficient on the debt equity ratio and the largest financial leverage risk
28
premium. This provides the biggest “bump” when a sample estimate is applied to the regulated
21
7.4%= 11%*0.577 + 5%(1-.5)*0.423
25
1
common equity ratio. In my example it would give an equity cost of 12.31%, 131 basis points higher
2
than the true equity cost and higher than using any other equity cost model that I am aware of.
3
Moreover the higher the market value and the bigger the need to cut the allowed ROE, with this
4
approach the higher the unnecessary financial leverage adjustment. It is not surprising therefore with
5
such high market to book ratios that the ATWACC approach is popular with utilities.
6
Q. DO YOU THINK THE MODELS ARE ACCURATE ENOUGH TO SET RATES?
7
A. No.To illustrate when the NEB set up its regulatory system in RH-2-94 several experts submitted
8
testimony on the impact of capital structure changes along the lines of the current testimony of Drs.
9
Kolbe and Vilbert. Dr. Sherwin and Ms. McShane, who at the time provided testimony on behalf of
10
the companies, concluded before the NEB (page 24 of the RH-2-94 decision)
11
12
13
14
“The finance models, even when adapted to the real world of Canadian utility regulation, cannot
provide the basis for determining a pipeline’s optimal capital structure.”
15
If they can’t determine the optimal capital structure then they can’t determine the leverage
16
adjustment. In that hearing Dr. Berkowitz and I also used models similar to those used by Dr. Kolbe
17
in this hearing, but expressed little support for them. As the Board noted in its Reasons for Decision
18
(page 24 of the RH-2-94 Decisison)
19
20
21
22
23
24
“Dr. Booth and Berkowitz concluded that these estimates are approximately the increases in
ROE required by investors. However, they noted the estimates are subject to error since they are
based on valuation formulas, which are as yet unproven. Moreover, they noted that these
formulas ignored the non-tax advantages of debt financing and the effects of financial distress.”
25
Finally, the NEB also noted Dr. Waters’ testimony (a frequent witness before the NEB at that time)
26
where he indicated that “To date empirical testing to more clearly describe the relationship (between
27
capital structure and the investors required return) has not been done successfully.”
28
29
The NEB’s summary from well over ten years ago is an accurate assessment of my views today and
30
it is still my judgment that the misgivings expressed by expert witnesses over ten years ago continue.
31
Despite the seeming precision of the estimates provided by Drs. Kolbe and Vilbert the estimates are
32
based on assumptions and models that have not been verified in the real world.
26