613
REGULATORY POLICY AND THE
PERFORMANCE OF ELECTRIC UTILITIES:
A SYSTEM DYNAMICS ANALYSIS
James M. Lyneis
Pugh-Roberts Associates, Inc.
Cambridge, Massachusetts
Dominic Geraghty
Electric Power Research Institute
Palo Alto, California
April 1983
1. DurRopucrton
The planning environment faced by the electric utility industry has
become increasingly complex. Not only have there been many ‘shocks’, such
as oil embargoes, escalating prices, and construction delays, but the
reactions of once almost-predictable and benign external entities have
become increasingly uncertain and disadvantageous to the utility. Con-
sumers have reacted strongly against price increases, such that load growth
has slowed dramatically and in sone instances, available capacity is higher
than that required from the standpoint of maintaining service reliability.
Investors have required greater returns as the financial performance of
utilities has fallen; and this has further weakened the utilities’ finan-
cial condition, Faced with consumer pressure, regulators have become more
reluctant to grant rate relief of sufficient magnitude to enable utilities
to earn their cost of capital. As a result, utilities are in poor finan-
cial condition, with falling bond ratings ani stock selling below book
value.
Managers and regulators are likely to have a difficult time correcting
these problems. This stems primarily from the interrelatedness of the
problem -- there is no single cause of the utilities plight. Rather, the
present situation has evolved over the last ten years as the individual
actors -- utility management, consumers, investors, and regulators -- each
has responded to the various shocks and the actions of the other actors in
such a way as to cause a generally deteriorating situation:
PUGH-ROBERTS ASSOCIATES. INC.
614
© management was unable to adjust long lead-time construc-
tion programs, resulting in unwanted capacity becoming
available in times of falling load growth;
© consumers responded to rapidly escalating rates by re~
ducing usage, but this increased fixed costs per kilowatt
hour further, creating upward pressures on rates;
© investors responded to inflation and the deteriorating
financial peformance of utilities by requiring a higher
risk premiun, but this further raised utility costs,
created upward pressure on rates, and worsened financial
performance; and
0 regulators have held down rates of return on utility
stock, thereby worsening financial performance and in-
creasing the cost of raising new equity and debt.
A downward spiral of higher costs, higher but inadequate rates, poor
financial performance, slower load growth, and even higher costs has deve-
loped from the combined actions of management, consumers, investors, and
regulators. As a result, there is likely to be no easy solution to the
utilities’ problems. lower inflation, higher (or lower) demand growth, or
more understanding investors and regulators alone will not dramaticelly
improve the situation. Rather, it will take a combined effort to reverse
the trends which have led to the present condition, and prepare utilities
for the plant expansion needed to meet demand growth in the nineties.
In response to the need for an integrated look at the problems of
electric utilities, Pugh-Roberts Associates, Inc. has developed a strategic
planning model for electric utilities. In various forms, it has been used
by utility industry investors, by individual utilities, and by research
organizations for analyzing alternative investment, management, and regula
tory strategies.
PUGH- ROBERTS ASSOCIATES. INC.
A wide range of policy issues can end have been explored with the
model. These issues include:
1. Capital investment policy
© reserve margin goals
© size of new plants constructed
© customer versus shareholders interests, when funds are
Limited
© investment in end-use and load management
control
© cancellation of plants under construction
0 diversification
2. Financial Policy
o dividend levels
© willingness to sell stock below book value
© debt levels
3. Regulatory Policy
© allowed rate of return
° WIP
© forward test year
This paper describes the use of the model to analyze the impact of
alternative regulatory policies on utility performance. As noted above
changes in regulatory policy alone will not solve all the problems of the
utilities. Nevertheless, as will be shown, regulatory policy does have a
strong impact on utility performance and must be a central element of any
strategy to revive the industry. The results of other analyses are re-
ported elsewhere (see 1, 2, 3 and 4). Finally, the reader is cautioned as
to interpretation of the results contained in this paper. These results are
PUGH-ROBERTS ASSOCIATES. INC.
615
intended to be representative of the electric utility industry. However,
individual utilities differ from the average, and the results may vary
depending on the assumptions used.
PUGH- ROBERTS ASSOCIATES. INC.
II. SPRUCTURE OF THE MODEL
Overview
The electric utility model is a behavioral simuletion model useful for
analysis of strategic and policy issues. ‘The model is behavioral in the
sense that it describes the cause and effect forces which determine the
behavior of the utility, for example the forces which lead to investment in
baseload capacity. The model is a simulation model in the sense that,
given the condition of the utility at 2 point in time (e.g. 1975), it
calculates the changes to that condition which result from the behavioral
or Ce ee ee
calculations are made every eighth of a year. So between say, 1975 and
2000, the model steps through, or simulates, 200 evolutions of the condi-~
tion of the utility.
The model is primarily useful for analysis of strategic and policy
issues because of its scope and level of detail. In scope, it consists of
a series of sectors representing the major activities of a utility and its
interaction with the external environment (customers, investors, regula-
tors, general economy). These sectors are:
1, Demand Generation
2. Capacity Planning
3. Power Generation
4. Financial Planning
5. Accounting
PUGH- ROBERTS ASSOCIATES. INC.
616
6. Capital Markets
7. Regulation
The model represents the activities within these sectors at a relatively
aggregate level of detail. It does not, for example, pinpoint the timings
and magnitude of security issues. It does identify the order of magnitude
of financing needs (+ 10 percent), and more importantly shows the impact of
alternative capital structures on utility performance. It provides quick
turnaround analyses (several hours to a couple of days) which consider the
entirety of the utility and its environment. These are the analyses useful
for evaluating strategy and policy questions.
Figure 1 highlights the key interactions among model sectors. An
aggregate demand for electricity is calculated in the Demand Sector, based
“on exogenously specified growth rates, and on the price of electricity.
Denand “drives” the Power Generation Sector and also is used as the basis
for load forecasting in the Capacity Planning Sector. Capacity is ordered
to meet the load forecast, subject to availability of funds. The Power
Generation Sector provides power in response to demand, within the con-
straints of capacity available. The Accounting Sector determines the
utility's financial performance, based on the amount of power delivered,
rates, and various categories of costs. The Financial Planning Sector
raises capital in response to the utility's financial performance and the
requirenents of the Capacity Planning Sector, and feeds information back
concerning availability of funds. The capital Markets Sector determines
the cost of debt and equity capital based on utility financial performance.
Finally, the Regulation Sector uses information about the utility's costs
and its rate base to establish an aggregate rate for all customers.
PUGH-ROBERTS ASSOCIATES. INC.
RATE
REGULATION
costs.
ACCOUNTING
RATE BASE
DELIVERED
POWER
DEMAND
GENERATION
[-
POWER
GENERATION
FINANCIAL,
PLANNING
FUNDS
AVAILABILITY OF
KEY INTERACTIONS AMONG MODEL SECTORS
CAPACITY
DEMAND FOR FUNDS:
HISTORICAL DEMAND
PLANNING
FIGURE |
CAPACITY
Figure 2 shows a more detailed representation of model sectors and the
major external inputs to the model. Demand is calculated for residential,
industrial and commercial users based on the number of customers in each
class (exogenously specified over time), a reference KWH per year for
customers in each class (exogenously specified over time), and an effect of
price on the KWH per year actually used. ‘he effect of price can be dif-
ferent for each customer class, depending on the real price of electricity
(adjusted for changes in real income), the short-term price elasticity, and
the long-term price elasticity.
Capacity planning and construction is modeled for four categories of
plant: peaking, ofl-fired, nuclear and coal. When lead-times and fin-
ancing permit, baseload rather than peaking units are constructed. The
fuel type of the baseload units is specified exogenously as a function of
time. Units constructed in the future are assumed to be coal-fired. Where
lead times do not permit construction of baseload units, peaking units are
constructed... This might occur when actual demand growth exceeds expect-
ations, or when financial constraints limit or delay construction of base-
load units.
‘The Finaticial Planning Sector of the model takes the demand for funds
and trys to raise debt, preferred, or common equity to meet any shortfall
not provided by retained earnings. The mix depends on the costs and avail-
ability of each type. When funds are not available, capacity construction
4s delayed or not started.
The Accounting Sector computes financial performance based on power
delivered, rates, and costs. Costs are computed for five distinct cate-
gories:
PUGH- ROBERTS ASSOCIATES. INC.
NUMBER OF CUSTOMERS
BY TYPE
7
aa
REFERENCE KWH PER
(7 CUSTOMER BY TYPE
EARNINGS
RETAINED
Coo oe st
MORE DETAILED REPRESENTATION OF MODEL
—--------- 4)
DEMAND FOR FUNDS
FIGURE 2
CAPACITY COST
BY TYPE
iL
Fuel
2. Operating and Maintenance
(Fixed and Variable Components)
3. Depreciation
4. Financing (Interest)
5. Taxes (General, Income)
Inflation rates, specified exogenously as a function of time, cause costs
to change from a reference level (e.g., capacity construction costs of
$1000/KW in 1980).
The Regulatory Sector computes only one aggregate rate per KWH. While
different rates might be computed for each custoner class, this aggregation
was felt appropriate given the assumption that rates are set by the same
regulatory body, and that they will be allocated proportionally (by cost of
service) to all customer classes.
In all, the model contains approximately 700 equations which describe
the sectors, their interactions, and the external environment.
Key Feedback Relationships
The model contains a large number of feedback relationships. The
key relationships, which involve price and demand, the capital markets, and
regulators, are described below.
Feedbacks Involving Price and Demand. The two feedback loops involv-
*
ing price and demand are shown in Figure 3. The first loop (solid lines)
* Arrows in the figure indicate the direction of causality between two
variables, while the sign at the end of the arrow indicates the “pola-
rity" of that causality. For example, an increase in cost per KWH in-
creases rates (+); an increase in rates, however, decreases demand (-).
PUGH- ROBERTS ASSOCIATES. INC.
619
12
DEMAND 4
-
Za
a
a
7
M2
CAPACITY + RATE
POWER +
DELIVERED
= COST PER
KWH
“4
i, -
FIGURE 3 PRICE FEEDBACK LOOPS
13
is a positive loop, and works as follows: given a certain amount of cap-
acity and therefore fixed costs, an increase in power delivered tends to
reduce cost per KWH. Other factors affecting rates being equal, rates are
also reduced, thereby increasing demand. The increase in demand further
increases power delivered, lowers costs, lowers rates, and so on. ‘This
loop is called @ positive feedback loop because it tends to feed on itself
in an ever-growing (or declining) spiral. (In the decline mode, positive
feedback loops are often called vicious circles).
The second loop shown in Figure 3 is a negative feedback loop which
acts to control the positive loop. As demand grows, more capacity is
needed to provide the power. As a result, fixed costs and cost per KWH
increase. In response, rates increase, thereby lowering demand. The loop
is negative in that an initial increase in demand stimulates actions which
decrease demand, whereas in the positive loop the increases feed on then-
selves.
The delays in the negative feedback loop are considerably longer than
those in the positive loop because of the length of time required to add
new capacity. ‘Thus, the positive loop can operate for a mumber of years
vefore the negative loop acts to control it.
Historically, the positive loop has been dominated by external trends:
rapid growth in population and standard of living have spurred on demand
growth, such that the perturbations caused by these feedback loops have not
been noticed. In fact, before the mid-seventies, the addition of capacity
tended to lower, rather than to raise, costs. Both loops, therefore, acted
to stimulate demand. Now, however, with slow demand growth, the effects of
these loops become important and can cause wide cycles on an underlying
PUGH- ROBERTS ASSOCIATES. INC.
14
growth trend, as will be described later in the discussion of the base case
aimulation.
Feedbacks Involving the Capital Markets. A second important set of
feedback loops involve the capital markets, as shown in Figure 4. The
first of these loops involves interest: an increase in interest rate tends
to raise interest charges; the increase in interest charges raises costs,
which in turn lowers interest coverage. ‘The reduction in interest cov-
erage, after a delay representing market perception and reaction to the
change, further increases the interest rate, initiating a downward spiral.
The interest loop is a positive feedback loop. ‘The decrease in interest
coverage initiates a second downward spiral; after a delay, it acts to
reduce stock price, which means that more shares must be issued to raise @
given volume of external funds. As more shares are issued, stock price
tends to fall further because of the dilution in earnings, and so on. Both
of the feedback loops further feed on themselves by reducing internal funds
generation and by raising external funds required: increasing interest
charges increases costs; increasing the number of shares raises dividend
payments.
The downward spirals produced by the positive feedback loops are
controlled by the negative loops shown by the dashed lines in Figure 5. As
interest coverage and stock price deteriorate, the utility either becomes
unable or unwilling to reise additional debt or equity. As a result,
available funds fall short of those required and the construction progran
is adjusted to reduce external funds requirements. Because less new stock
and debt are issued, the downward spirals are slowed or eliminated (until
service and reliability criteria force construction).
PUGH- ROBERTS ASSOCIATES. INC.
620
15
CONSTRUCTION
PROGRAM
+ it a. a,
+ FUNDS
SHARES: ‘
ene |
GROWTH IN
EARNINGS PER
SHARE
costs:
STOCK INTEREST
BRick SL" COVERAGE
INTEREST
RATE
INTEREST.
CHARGES
FIGURE 4 CAPITAL MARKET FEEDBACKS
—
DEET
621
While these capital market feedback loops have existed in the past, it
+ is unclear how important they have been. As will be seen in the discussion
_ — — AVAILABLE
aa roe of the Base Case, the conditions which initiate the downward spirals are
== /
ge! Pa , ;
co constauction. likely to exist in the later eighties and early nineties.
- PROGRAM \
- \ Feedbacks Involving Regulators. An important agent in many of these
-
yf / \ feedback loops is the regulatory authority. ‘The authority takes inform
7 . EXTERNAL
OS
ation about costs and rate base and converts then into rates. Changes in
is REQUIRED.
a DIVIDENDS ———" 5 : rates, as noted above, feed back to influence costs via demand and capacity
| expansion. The model represents the rate-setting process as it is in real
GROWTH IN | life -- imperfect. There is assumed to be a lag of one year between the
EARNINGS PER
SHARE | time a rate case is filed and finally approved, and significant delays in
N
cosTs+ | responding to change in the inflation rate.
stock + INTEREST. | An important set of feedback loops connect the utility, the regula-
PRICE S—__—"_ COVERAGE
pest | tore, and the capital markets, as shown in Figure 6. If, for whatever
| - / reason, interest coverage should fall, interest rates increase as described
| INTEREST /
RATE / above. After regulatory delays, the increase in interest rate leads to an
\ i increase in rates, which in turn improves earnings and interest coverage,
NG =p INTEREST. 7 thereby halting a further increase in interest rates. In other words, as
. 7
Sa, _- the risk to debt holders increases, the cost of this to the utility is
~~
reflected in rates. A similar feedback through risk to equity holders is
also included in the model, but only takes effect for values of interest
coverage below 3.0.
Other Feedback Relationships. As noted previously, there are many
FIGURE 5 CAPITAL MARKET FEEDBACKS WITH feedback loops in the model beyond those given in Figures 3, 4, and 5. An
NEGATIVE CONTROL LOOPS
example of some of the more subtle feedback loops involving price are shown
in Figure 7. These loops show how efforts to hold down rates in the
short-term can increase them in the long-term. Historically, regulatory
? PUGH- ROBERTS ASSOCIATES. INC.
622
INTEREST — —
COVERAGE
aN
\
\
AvareaeLe
EARNINGS INTEREST \ INFLATION 2 .
} ee
I RETURN:
_ INELATION:
RATES ATE 7 . :
BASE-LOAD
7 é CAPACITY Aone
oe 4 E. .
BLLOMED | li 7 PEAKING
ON EQU CAPACITY aie
= RATE <>
FUEL COSTS PER
+ KWH
FIGURE 6 FEEDBACKS INVOLVING REGULATION
FIGURE .7 MORE SUBTLE FEEDBACK LOOPS INVOLVING PRICE
20
bodies have been slow to increase allowed return in response to changes in
infletion rate. While this tends to slow the rate of increase in rates, it
also reduces funds available for construction in two ways: first, by
directly reducing internal cash flow; and second, by reducing the real
return to investors, and/or increasing the risk of those investments,
thereby making external financing more difficult to obtain and/or more
costly. As a result, shortages of funds constrain the construction of
base-load units. Should demand continue to grow (as it is likely to do
with the added stimulus from the price feedback loops), peaking capacity
will eventually be needed to meet demand. But the use of peaking capacity
for baseloads raises fuel costs per KWH, thereby raising rates over what
they might have been, had base-load units been constructed. The model
contains many feedback relationships of this type.
Important Model Assumptions
In addition to a structure which states how the pieces of the utility
and the environment interact, the model contains assunptions about external
trends, the strength of reactions by external agents (e.g. capital markets
and regulators), and management policies. The structure together with the
assumptions determine the time behaviour of variables in the model. Impor-
tant assumptions are noted below. ‘These assumptions are meant to be rea-
sonable and representative of the utility industry. They constitute a
hypothetical electric utility.
External Trends. Assumptions regarding external trends fall into two
categories: (1) factors affecting demand growth, and (2) cost inflation
rates. Specific assumptions used in the Base Case simulation of the model
PUGH-ROBERTS ASSOCIATES, INC.
623
21
are given in Table 1. As can be seen in the table, demand growth, exclu-
sive of price changes, is expected to average 2.3 percent per year. Price
changes work through short- and long-term elasticities (additive effects)
to change demand growth fron the rates given above. Inflation in utility
costs is assumed to exceed general inflation rates.
Reactions By External Agents. External agents determine three factors
of importance to utilities: interest rates, stock price, and rates. How
each is modeled is discussed below.
The interest rate on new debt NINTR equals the sum of three compon-
ents: a risk-free rate RFINT, inflation preniun IFPD, and a risk-premium
RFD:
WINTR=RPINT+IFPD*RPD
NINTR - New Interest Rate (fraction/year)
IFPD - Inflation Premium for Debt (fraction/year)
RFINT - Risk-free Interest Rate (fraction/year)
RPD = Risk Premium for Debt (fraction/year)
The risk-free rate is assumed to equal @ constant 2.5 percent; the infla-
tion premium is simply a one-year average of the inflation rate. The risk
premiun for debt is modeled as a function of interest coverage, since
interest coverage is a key factor in utility bond ratings, and is also a
reasonable proxy for other risk indicators. ‘The risk-prenium is assumed to
rise steeply as interest coverage falls. ‘The interest coverage used is a
weighted average of coverage including and excluding allowance for funds
used during construction (AFUDC).
Investors in utility stock are assumed to value it much like debt,
that is, by dividend yield. As indicated in the equations below, market
price per share MPS equals dividends per share DIVPS divided by net stock
discount rate NSDR, where NSDR equals the sum of a risk-free interest rate
PUGH- ROBERTS ASSOCIATES. INC.
22
RFINT, an inflation premium IFPD (same premium as for debt), a risk-premium
for equity RPE, and the negative of anticipated growth in dividends per
share AGDPS.
MPS = DIVPS/NSDR
NSDR = RFINT+IFPD+RPE-AGDPS
MPS = - Market Price Per Share ($/share)
DIVPS - Dividends Per Share ($/year/share)
NSDR - Net Stock Discount Rate (fraction/year)
RFINT - Risk-Free Interest Rate (fraction/year)
IfPD - Inflation Premium for Debt (fraction/year)
RPE - Risk-premium for Equity (fraction/year)
AGDPS - Anticipated Growth in Dividends Per Shere (frac-
tion/year)
‘The risk-free rate and inflation premium of debt are the seme es that used
in determining new interest rate.
‘The risk premium of equity is a function of interest coverage (same
coverage as for risk premium of debt). Given that most utility stockhol-
ders view their stocks as near-debt, interest coverage is a reasonable
indicator of the risk of being paid dividends. In the model, risk-premium
rises steeply when interest coverage falls.
Anticipated growth in dividends per share is based on historical
dividend growth. Anticipated growth is assumed to equal historical rates
of growth, as calculated by the model, over the last several years. The
higher the growth rate, the lower the discount rate. The above stock
valuation model gives a good fit to the historical stock prices of many
utilities modeled in earlier work.
The rate set by the regulatory body for this hypothetical utility is
the sum of three components: (1) fuel cost adjustment; (2) other costs;
and (3) return on rate base. Changes in fuel costs are passed through with
PUGH-ROBERTS ASSOCIATES. INC.
624
23
TABLE 1
BASE CASE ASSUMPTIONS: EXTERNAL ENVIRONMENT
1, Demand Growth Rates Assuming Constant Real Prices and Real Income: 2.3%
per annum
2. Price Elasticity: -1.0 for all customer classes
3. Inflation and Real Income:
General Inflation Rate of 8% p.a. (actual CPI used 1980,1981)
Increment in Utility Costs from General Rate -
Increment Increment
1983-1990 After 1990
Capacity Cost +n +z
081 Cost 41.5% 41.5%
Nuclear Fuel Cost 4% -1%
Coal Cost 40.5% 40.5%
‘08M Cost +12 41%
General Taxes +18 +n
4, Regulation: @ Assuming stable inflation rate, regulators will allow
a real return on equity consistent with risk level by
1990 (assumed to be 8%).
@ Regulatory Delay of 1 year.
@ No forward test year or CHIP.
24
a three-month lag; the latter two components must be approved in a reg-
ulatory proceeding. The delay in granting a new rate is set at a constant
one year. There is no forward test year nor CWIP allowed.
The allowed rate of return is the sum of the allowed debt, preferred,
and equity returns, weighted by their percentage of the capital structure.
Debt and preferred returns are based on actual charges paid; the allowed
return on equity is the sum of a real return and an inflation adjustment,
which is a function of a five-year average of the inflation rate. Histori-
cally, allowed returns on equity have not kept pace with inflation such
that real returns have fallen. ‘There are two possible explanations for
this: (1) regulators have been slow in recognizing the permanence of high
rates of inflation; and (2) regulators have responded to consumer pressures
and allowed real returns to fall, even though they accept the high infle-
tion rates as "permanent". Either way, the model represents both of these
explanations by basing the allowed return on equity on an average of infle~
tion.
Management Policy Variables. The two important areas of management
policy relevent to a strategic planning model are capacity expansion policy
and financing policy. Table 2 lists the key elements of the hypothetical
utility's policies. Policies in the model state how utility management
(and for that matter investors and regulators) respond to changing condi-~
tions. They are an integral part of the feedback structure of the model.
PUGH- ROBERTS ASSOCIATES. INC.
625
25
TABLE 2
BASE CASE ASSUMPTIONS: UTILITY POLICIES
1. Capacity -
@ Desired Reserve Margin - 20%
@ Desired Fuel Type - Only new coal plants after present
construction, except for normal
amounts of peaking capacity.
@ Construction Lead Times - 8 years for coal Baseload
3 years for peaking
© No significant investment in conservation or load
management.
2. Financing
@ Desired Capital Structure - 50% Debt, 38% Common,
12% Preferred
@ Dividend Payout Objective - 75%
PUGH- ROBERTS ASSOCIATES. INC.
26
III, BASE CASE PERFORMANCE OF HYPOTHETICAL UTILIITY
The Base Case is simply a simulation from 1975 to 2000 which is pro-
duced assuming a continuation of present management and regulatory prac-
tices, and likely assumptions regarding the evolution of external trends.
It is a look at the historical and likely future performance of a hypo-
thetical electric utility.
Goal Base Case.
Figures 8 through 15 show the projected Base Case performance of the
hypothetical coal-based electric utility from 1980 to 2005. Figure 8 shows
the trends in capacity, peak load, power delivered, and reserve margin.
Time runs across the bottom axis; the scales for the variables plotted are
given along the vertical axis (in the scales, "I" stands for thousands, "M"
for millions, and "B" for billions).
‘From 1980 to 2005 demand growth averages 2.2 percent, slightly below
the growth rate of 2.3 yeroeat tn customers and usage per customer asouning
constant prices. However, the rate of growth over the period is not at all
smooth, whereas customer growth is. Both of these deviations are caused by
variations in real price, as shown in Figure 9.
In the early eighties, prices are high relative to the levels of the
seventies. Hence, price feedback effects on consumption are keeping demand
nearly constant in spite of the 2.3 percent per year growth in customers
and usage per customer assuming constant real prices. Reserve margins
therefore continue to grow over the 1980 to 1985 period as construction
PUGH-ROBERTS ASSOCIATES, INC.
2005.
2000,
-———,, oe
1995,
TIME
FIGURE &
DEMAND AND CAPACITY
1990,
—— — DEMAND (KWH/YR) (0., 180, 8)
20, M)
1985.
CAPACITY (KW) (O.
————PEAK DEMAND (KW) (0,, 20.)
190.8 RESERVE CAPACITY (FRACTION) (0... 4)
4
20.
10.
2005.
2000.
1995.
TIME
FIGURE 9
1990,
RATE ($1982) (40. A, 80. A)
1985.
RATE AND CAPITAL EXPENDITURES IN CONSTANT 1982 DOLLARS
—— _—— CAPITAL EXPENDITURES ($1982) (0. , 2000.
1980.
80.4
70.4
1500, M
60. A
1000. M
S0.A
500. M
40.4
2000. M
29
started earlier is completed. Prices are high for two reasons: first,
increases in fuel costs above the rate of inflation; and second, increases
in fixed charges per KWH at the high reserve margins mean that the costs of
unused capacity must be spread across a smaller base of power delivered.
As the utility perceives the slow load growth and high reserve margin,
the construction program is reduced (see Figure 10): in 1980, 3 plants
were under construction; by 1985 only one is. This low level of construc
tion is maintained until 1988.
With the reduced construction program, real prices are nearly flat
from 1980 to 1989. Demand growth then acts to drive down reserve margins;
as a result fixed charges are spread over a larger KWH base. A fall in
real prices then further stimulates demand growth--between 1987 and 1995
demand growth exceeds the rate inherent in customer growth.
In response to the renewed load growth and declining reserve margins,
the utility once again gears up the construction program. But the unanti-
cipated growth stimulated by falling real prices causes reserve margins to
fall below the utility's 20 percent objective because of the long lead
times of base load plants. Peaking units are added, and margins improve in
the late nineties.
The cost of fuel for the peaking units begins to drive up real prices.
Increases in utility costs ebove general inflation and the arrival of new
baseload units into the rate base continue the upward trend. As a result,
price feedback effects cause demand to reach a peak and level off between
2000 and 2005 such that reserve margins rise above 20 percent.
PUGH- ROBERTS ASSOCIATES. INC.
FIGURE 10
CAPACITY ON-LINE AND UNDER CONSTRUCTION
BASELOAD CAPACITY UNDER CONSTRUCTION (KW) (O., 6400. T)
—— —— — BASELOAD CAPACITY (KW) (0., 20.M)
——-——~ PEAKING CAPACITY (KW) (0., 20. M)
30
ad ~
wore
2005.
2000.
—- =H
a
—-—.
1995.
a
1990.
1985,
6400. T
20.M
4800. T
15.
1600, T
TIME
628
31
The financiel performance of the utility is shown in the remaining
figures. Figure 11 gives the behavior of return on equity and return on
rate base. By assumption, the allowed return on equity increases to 16
percent (8 percent inflation plus 8 percent real return.) Realized re-
turns, however, except for a brief period near 1990, average well below
allowed. Returns fluctuate because of the discrete nature of rate cases:
they rise immediately after a rate case, then deteriorate as inflation
drives up costs while rates remain constant (except for fuel charge pass-
throughs), and as new plants are brought into service but regulatory lag
delays their inclusion in the rate base. Returns are therefore the lowest
when inflation is highest and when the most construction is occurring.
Conversely, returns increase toward allowed when costs are falling and the
construction progren is low (as in the late eighties).
Figure 12 shows sone ‘per share’ data for the utility. Earnings and
dividends per share grow at rates averaging 4 percent per year from 1962 to
1992 as the reduced construction program, relative to internal funds flow,
obviates the need for new equity, falling reserve margins improve equity
returns, and then as AFUDC begins to grow again after 1987. But funding
construction in the nineties requires new equity and flattens earnings and
dividend growth, particularly when regulatory lag causes delays in convert~
ing AFUDC to return on rate base and expenses to revenue. During the
eighties, AFUDC percentage of earnings falls to 25 percent, but rises to
high levels in the late nineties because of the very high levels of con-
struction work in progress relative to present assets.
For a brief time near 1990, market price per share equals book value
per share (Figure 13). A reduction in interest rates, dividend growth, and
a reduction in risk premium all act to stimulate market price. The
PUGH- ROBERTS ASSOCIATES. INC.
FIGURE 11
RETURN ON EQUITY
RETURN ON EQUITY (0.,.2) ——--———- ALLOWED RETURN ON EQUITY (G.,.2)
———_———_RETURN ON RATE BASE (0.,.2)
+2
ooo —_ TT
215) =
a
aT
05
1980. 1985. 1990, 1995, 2000, 2005.
TIME
FIGURE 12
COMMON STOCK DATA
EARNINGS PER SHARE (0., 16.)
—— —— DIVIDENDS PER SHARE (O., 16.)
—— —— DIVIDEND PAYOUT RATE (O.,1.)
‘NUMBER OF SHARES (0., 400. M)
18 "~~ AFUDC PERCENTAGE OF EARNINGS (0., 1.)
119 1980. 1985. 1990, 1995, 2000. 2005,
‘0 IME
we
€€
629
FIGURE 13
MARKET .PRICE PER SHARE
RISK-FREE RATE (0.,. 16)
‘MARKET PRICE PER SHARE (0., 120.)
a — — BOOK VALUE PER SHARE (0., 120.)
——_———_NET DISCOUNT RATE (0... 16)
-~~RISK PREMIUM OF EQUITY (O,,. 16)
120.
16)
~~~ -GROWTH ADJUSTMENT (O,
34
, é
| a oR
fo oh \A
oi }
il,
! re
nn !
i i /
| if 1 / Bg
\ fF
N |! Ko of
iy [t
tig
ral
i
3
630
35
reduction in risk occurs because, with the reduced construction, interest
coverage improves dramatically (see Figure 14) and earnings quality im-
proves (reduction in AFUDC percentage, Figure 12). But the improvement is
short-lived. Once construction resunes, a downward capital markets spiral
causes financial performance to rapidly deteriorate: financing construct-
ion reduces interest coverage and earnings growth; as these fall the cost
of additional financing increases; performance further deteriorates with
the next round of financing. The spiral is broken only when performance
falls to such a low level that baseload construction outlays must be limit-
ed. As a result, the utility in 2005 is using more than ‘normal’ peaking
generation.
Finally, Figure 15 shows the capital structure of the utility and
internel financing. During the late eighties, internal funds are nearly
sufficient to cover all construction expenditures. No new debt is issued
except that to cover retirements. Extra growth in retained earnings re-
duces debt percentage of total capital. But once construction resumes,
internal financing adequacy falls sharply.
PUGH- ROBERTS ASSOCIATES. INC.
FIGURE 14
FINANCIAL PERFORMANCE
TIMES INTEREST EARNED INCL. AFUOC (1.5, 4.5)
———— TIMES INTEREST EARNED EXCL. AFUDC (1.5, 4,5)
MARKET PRICE TO BOOK VALUE RATIO (0., 2.)
-NEW INTEREST RATE (0... 2)
—— _—— EMBEDDED COST OF DEBT (0... 2)
Lee
_~" 1)
.0
19 1980. 1985. 1990, 1995. 2000. 2005.
TIME
FIGURE 15
CAPITALIZATION PERCENTAGES
DEBT PERCENTAGE OF CAPITALIZATION (O., 1.)
— ——— PREFERRED PERCENTAGE OF CAPITALIZATION (O., 1.)
= —---—= COMMON PERCENTAGE OF CAPITALIZTION (0.,1.)
1 ———- —— ~ PERCENTAGE INTERNAL FINANCING (0., 1.)
—-
/ \
+75 4
/ \
Lif | ee
i
-Oyo80. 1985, 1990. 1995, 2000, 2005,
TIME
9
Teo
“ec
38
IV, REGULATORY POLICY AND UTILITY PERFORMANCE
A number of regulatory policy changes have been proposed to improve the
financial condition of electric utilities. These changes include:
1. Allowing a Higher Return on Equity -- Regulators have been reluc-
tant to allow the real return on equity to keep pace with infla-
tion and increased financial risk. In the late sixties and early
seventies, when inflation was low, many utilities were allowed a
real return on equity near 9 percent. Today, that real return is
vetween 4 and 6 percent, depending on one's perception of the
underlying rate of inflation. As inflation stabilizes, the real
return will move toward 8 percent in the Base Case. What would
be the consequences if regulators allowed the utilities a 9
percent real return on equity, approximately the level histori-
cally allowed in low-inflation periods?
2. Reducing the delay between filing and granting of rates to 3
months from 1 year in Base Cas
3. Calculating rates based on expected costs and rate base in a
“forward test year.”
4. Allowing construction-work-in-progress (CWIP) to be included in
the rate base (with full AFUDC between rate cases) -- For most
utilities, a return is allowed only on producing electric plant,
and not on construction, on the argument that it is not fair to
charge today's customers for construction needed for tomorrow's
customers. But including CWIP should improve the utilities cash
flow and improve financial performance.
5. “Nearly perfect regulation” -- higher return, CWIP, and instan-
taneous pass-through of cost and rate base changes.
A primary objective of all these policies is to raise the earned return
on equity. Figure 16 shows how the policies perform in this regard. Allowing
a higher rate of return has a surprisingly small effect, less than the full
percentage point increase indicated by the policy change. This occurs be-
cause, with higher allowed return, financial performance improves somewhat and
PUGH-ROBERTS ASSOCIATES. INC.
632
39
relaxes Base Case pressures, which were already allowing more than the 8
percent normal return. A similar relaxation occurs in the other policies.
Note that as regulatory lag is progressively relieved through shorter delay,
forward test year, and imediate pass-through, return on equity increases.
GWIP also improves return on equity because it reduces the strength of the
capital market feedbacks, which were acting to ‘inflate’ interest costs and
thereby drive down allowed return given regulatory lag.
As return on equity improves, so do the other measures of financial
performance. Figures 17, 18, 19 and 20 compare the behavior of earnings per
share, interest coverage, market price per share and market-price-to-book-
value ratio, respectively.. Table 3 provides a numerical comparison. ‘The
improvement in these other financial measures are consistent with the improve-
ment seen in return on equity.
Note that market price to book value ratio remains above 1.0, and earn-
ings per share growth keeps pace with inflation, only with nearly perfect
regulation (although forward test year comes close). The other policy changes
improve performance over Base Case levels, but reasonable performance cannot
be sustained once construction expenditures resume. The need to finance these
expenditures at inadequate rates of return drive financial performance down to
minimum acceptable levels via the capital markets feedback loops. Note that
with policies which improve financial performance, capital expenditures are
actually higher than Base Case levels in spite of lower demand. In the Base
Case, construction expenditures are being held down below planned levels
because of poor financial performance--peaking capacity is being added and
baseload construction needed after 2005 being deferred. As financial per-
formance improves, expenditures increase and drive down performance.
PUGH- ROBERTS ASSOCIATES. INC.
FIGURE 16
RETURN ON EQUITY -- REGULATORY POLICIES
BASE (0... 2)
HIGHER RETURN (0.,..2)
= CWIP (O.,.2)
“FORWARD TEST YEAR (0.,.2)
~NEARLY PERFECT REGULATION (0.,.2)
7 SHORTER REGULATORY DELAY (O.,.2)
ai
“15 = Feet eryer tte.
= _
OF LON
> ar
o PLAIN
ol,
05
if 1980. 1985. 1990, 1995. 2000. 2005.
TIME
FIGURE 17
EARNINGS PER SHARE -- REGULATORY POLICIES
BASE (0., 40.)
——— ~~ HIGHER RETURN (0., 40.)
———— CHIP ., 40.)
~-FORWARD TEST YEAR (0., 40.)
——~--——- NEARLY PERFECT REGULATION (0., 40.)
77 SHORTER REGULATORY DELAY (0., 40.)
40.
"1980. 1985. 1990. 1995. 2000. 2005.
TIME
oy
Wy
€€9
FIGURE 18
INTEREST COVERAGE -- REGULATORY POLICIES
BASE (1.5, 4.5)
HIGHER RETURN (1.5, 4.5)
WIP (1.5, 4,5)
‘ORWARD TEST YEAR (1.5, 4.5)
——--——- NEARLY PERFECT REGULATION (1.5, 4.5)
~~~ SHORTER REGULATORY DELAY (1.5, 4.5)
4S,
a™,
toe O™.
ws =a eer ty ",
F: a ae
a ee ard Sa ae
eo.
NANO FR,
3. Nae Bice”
\ hat |
™
_— ra
2.25
1.5)
1980. 1985. 1990, 1995. 2000. 2005,
TIME
FIGURE 19
MARKET PRICE PER SHARE -~ REGULATORY POLICIES
BASE (0. , 200.)
HIGHER RETURN (0. , 200.)
— NEARLY PERFECT REGULATION’ (0. , 200.)
——--—— ~~ SHORTER Rt -ATORY \Y 0. , 200.
a EGULATORY DELAY (0. , 200. )
150.
1980. 1985, 1990. 1995. 2000.
TIME
“2005.
a
ey
veo
44
CWIP helps somewhat, but in between rate cases, AFUDC lowers the quality of
reported earnings. With regulatory lag, a full cash return on construction is
not realized.
Improvement in financial performance is not without cost to consumers
(see Figure 21). In fact, to achieve sustained reasonable performance via
nearly perfect regulation requires substantially higher. prices. It is likely
that some combination of changed utility policies and improved regulation can
achieve sustained reasonable performance at less cost to consumers.
Finally, Figures 22 and 23 show the consequences of these policy changes
for demand and reserve margin.
PUGH-ROBERTS ASSOCIATES. INC.
635
FIGURE 20
MARKET PRICE-BOOK VALUE RATIO -- REGULATORY POLICIES
BASE (0.,2.)
—— —— HIGHER RETURN (0.,2.)
aaa CHIP O22
------ FORWARD TEST YEAR (0.,2.)
——-——--NEARLY PERFECT REGULATION (0.,2.)
——--——~-~SHORTER REGULATORY DELAY (0.,2.)
&
_—-
=4
ae
ee |
2.
2005.
2000.
1995.
TIME
1g90,
1985,
1960.
TABLE 3
COMPARISON OF ALTERNATIVE REGULATORY POLICIES, 1982- 2005
Higher Shorter Forward Nearly Perfect
Base Return Delay Test Year cwIP Regulation
Growth in Power
Delivered 2.28/year = 1.9%/year =—1.8%/year = «1.5%/year ——1.5/year 1.0%/year
Growth in Earnings
Per Share 1.8%/year —«3.4%/year 4. 1/year 6.34 /year 3. 88/year 7.3t/year
Total Capital
Expenditures 18.7 22.6 25.2 24.3 24.0 21.0
(Billion 1982 $)
Total Net Income
ts Commi 1.9 91 10.4 heey 9.3 9.7
(Billion 1982 $)
Total Revenue
Requirements 74.6 75.3 75.2 75.0 74.9 72.0
(Billion 1982 $)
FIGURE 21
PRICE (1982$) -- REGULATORY POLICIES
BASE (40. A, 80. A)
—— —— HIGHER RETURN (40. A, 80. A)
————CWIP (40, A, 80. A)
~--=--FORWARO TEST YEAR (40. A, 80. A)
NEARLY PERFECT REGULATION. (40. A, 80. A)
SHORTER REGULATORY DELAY (40. A, 80. A)
BOA
40.A
1980. 1985. 1990,
TIME
1995.
2000, 2005.
oy
Ly
9¢9
FIGURE 22
POWER DELIVERED -- REGULATORY POLICIES
BASE (35, B, 75. 8)
—— — HIGHER RETURN (35. B, 75. 8)
——— WIP 35. 8, 75. 8)
FORWARD TEST YEAR (35, B, 75. 8)
NEARLY PERFECT REGULATION (35. B, 75. 8)
SHORTER REGULATORY DELAY (35. B, 75. 8)
FIGURE 23
RESERVE MARGIN -- REGULATORY POLICIES.
————BASE_(0.,.4)
—— —— HIGHER RETURN (.,.4)
—CWIP (0.,.4)
FORWARD TEST YEAR (0.,. 4)
NEARLY PERFECT REGULATION (0... 4)
SHORTER REGULATORY DELAY (0.,. 4)
TIME
*Oyo80. 1985. 1990. 1995. 2000,
2005.
ay
6”
Le9
50
+ CONCLUSIONS
The performance of the utility is strongly affected by regulatory policy.
Most actions which significantly improve financial performance are in the
regulatory area. Increased rates, however, need not be counter to the inter-
ests of the utility's customers. In fact, ‘an adequately funded baseload
construction program should in the long-run (here beyond 2005) provide better
service and lower rates by reducing reliance on purchased power and peaking
units. Further, we have assumed here that utility management will continue to
invest in spite of returns below the cost of capital. Should this not be the
case, the benefits of improved regulation become more pronounced.
Improved regulation in combination with actions by the utility, can
improve performance at little cost to consumers (based on additional simule-
tions not included here). Examples of such actions are: (1) the installation
of end-use and load management controls in the 1990's to reduce demand, im-
prove reserve margins, and reduce construction needs; and (2) less reliance on
debt to improve interest coverage, thereby reducing interest costs and raising
stock price. By themselves, these actions tend to reduce prices, and there-
fore offset the price increases accompanying improved regulation.
PUGH-ROBERTS ASSOCIATES. INC.
638
51
REFERENCES
Geraghty, D., “Utility Corporate Models Applied to the End-Use Versus
Supply Decision", Public Utilities Fortnightly, Sept. 10, 1981.
Geraghty, D. and J. Lyneis, “Feedback Loops: The Effect of External
Agents on Utility Performance", Public Utilities Fortnightly, Sept.
2, 1982.
Iyneis, J. and D. Geraghty, “An Electric Utility Strategic Model Based
on System Dynamics", Proc. 10th World Congress on Systems Simulation
and Scientific Computation, Montreal, Canada, August 1962.
Geraghty, D. and J. lyneis, "Utility Investment Strategies: Reconci-
ling the Objectives of Different Stakeholders", Proceedings of the
IEEE Power Engineering Society, January 1983.
PUGH- ROBERTS ASSOCIATES. INC.
CC BY-NC-SA 4.0