ELECTRIC UTILITIES' ACCOUNTING FOR CONSTRUCTION
WORK IN PROGRESS: THE EFFECTS OF ALTERNATIVE
METHODS ON THE FINANCIAL STATEMENTS, UTILITY
RATES AND MARKET TO BOOK RATIO












BY

GERALDINE WESTMORELAND


A DISSERTATION PRESENTED TO THE GRADUATE COUNCIL
OF THE UNIVERSITY OF FLORIDA IN
PARTIAL FULFILLMENT OF THE REQUIREMENTS
FOR THE DEGREE OF DOCTOR OF PHILOSOPHY



UNIVERSITY OF FLORIDA

1979















ACKNOWLEDGMENTS


A most respectful and sincere thanks is extended to

Dr. Eugene Brigham who has provided the encouragement and

guidance necessary for this dissertation. The Public

Utility Research Center and its members are gratefully

acknowledged for their financial support and the facilities

made available. I am most grateful to the members of my

supervisory committee, Dr. E. Daniel Smith, Dr. Sanford V.

Berg, and Dr. Eugene F. Brigham for their time and support.

A special note of thanks goes to Mr. William Talbot of the

Florida Public Service Commission and his accounting staff

for their capable assistance. I would also like to

acknowledge Katherine Williams for her excellent typing and

thoughtful assistance. My acknowledgments would be incom-

plete without special thanks to my friend Calvin Padgett,

who has contributed much encouragement.
















TABLE OF CONTENTS


ACKNOWLEDGMENTS . . . . . . . . .

LIST OF TABLES . . . . . . . . . .

LIST OF FIGURES . . . . . . . . .


Page



vii

xi

xlii


CHAPTER I ELECTRIC UTILITIES' ACCOUNTING FOR
CONSTRUCTION WORK IN PROGRESS: THE
EFFECTS OF ALTERNATIVE METHODS ON
THE FINANCIAL STATEMENTS, UTILITY
RATES AND MARKET TO BOOK RATIO . . .

Motivation for the Study . . . . . .
Statement of the Problem . . . . . .
Methodology . . . . . . . . . .

CHAPTER II PRIOR RESEARCH . . . . . . .

Accounting Literature on Theory of
Capitalization . . . . . . . . .
Electric Utilities' Accounting for CWIP . . .

The Quick Study . . . . . . .
The Arthur Andersen Study . . . . .
Pomerantz and Suelflow Study . . . .

Cost of Capital Studies Relevant to CWIP
Accounting . . . . . . . . . .

CHAPTER III REGULATORY THEORY OF ELECTRIC UTILITIES'
CWIP ACCOUNTING PRACTICES . . . .

Introduction . . . . . . . . .
Regulatory Theory . . . . . . .
Microeconomic Prescriptions . . . . . .
Legal Precedents and Embedded Cost Approach . .
Effect of Rate Base on Utility Rates . . .
Effect on Utility Rates Caused by Rate
Base Method . . . . . . . . .


xiii











Effect on Utility Rates Caused by Capitaliza-
tion of AFUDC with no CWIP in the Rate Base 46
Model of Hybrid Method: Inclusion of CWIP in
Rate Base with Offsetting AFUDC Credits . . 46
Effect of AFUDC on Interest Coverage Ratio . 51
Market to Book Ratio as a Cost of Capital
Indicator . . . . . . . . ... 53

CHAPTER IV COMPUTER SIMULATION . . . . .. 56

Overview of the Model . . . . . ... 56
Model Inputs and General Assumptions ... . 56
Model Outputs . . . . . . . ... 58
Logic of Projection . . . . . ... 60
Results of Computer Simulation . . . ... 61

Effect on Rate Base . . . . . .. 61
Effect on Utility Rates . . . . .. 63
Effects on Operating Income and Net Income. 66
Effects on Cash Flows after Dividends . 70
Effect on Internally Generated Funds as a
Percentage of Annual Investment in CWIP 73
Effect on Times Interest Earned Ratio (TIE) 78
Effect on Not Compounding AFUDC . . .. 82
Common Dividends Paid as a Percentage of
Net Income Available for Common Exclud-
ing AFUDC . . . . . . ... 34
Effect of AFUDC on Normalized Accounting. 86
Effect of Regulatory Lag on Simulation
Results . . . . . . . ... 87

Summary and Conclusion . . . . . ... 88

CHAPTER V THE EFFECT OF AFUDC ON THE MARKET TO BOOK
RATIO AND THE PRICE TO EARNINGS RATIO
OF ELECTRIC UTILITIES . . . . .. 92

Regression Models . . . . . . ... 92

Basic Assumptions and Tests for Multiple
Regression Models . . . . ... 95
Model Variables . . . . . ... 95

Hypotheses on the Effect of AFUDC on the
Market to Book Ratio . . . . . .. 97

Hypothesis 1 . . . . . . ... 98
Hypothesis 2 . . . . . . .. .104
Hypothesis 3 . . . . . . ... 109









Page

Price/Earnings Ratio Trend . . . . .. 116

Test . . . . . . . . .. 116
Results . . . . . . . ... 116

Summary . . . . . . . .... . .122

CHAPTER VI CURRENT ELECTRIC UTILITY PRACTICES . 124

CWIP Accounting Methods . . . . . .. 125

The Capitalization of AFUDC . . . .. .125
CWIP in Rate Base with Offsetting AFUDC
Credits to Revenue Requirements ... .125
CWIP in the Rate Base . . . . .. 126
Combinations . . . . . . .. 127

Implementation of Methods Which Capitalize AFUDC 127

Time Period and Expenditure Base ... .127
AFUDC Rate . . . . . . . 131
Accounting System and Financial Reporting 132
Interest Tax Benefits . . . . .. 139
The Effects of Different Asset Valuation
Bases on Cumulative Deferred Taxes . . 144

Industry and Commission Surveys . . . .. .148

CWIP Accounting Practices . . . .. .149
Methods Used to Determine Amount of CWIP
to Include in Rate Base . . . ... .152

Evaluation of financial integrity .154
Ultimate consumer as a basis for
allocation (Intergenerational Equity) 156
General approaches used by selected
states . . . . . . . 158

Interest Tax Benefits Practices . . .. .159
Compounding of AFUDC . . . . .. 166
Effects of AFUQC Earnings on the Ability
to Finance . . . . . . . .. 171
Projected Importance of CWIP Accounting 174
Comparison of Utilities Practices and
Regulatory Policy . . . . . .. 176
Comparison of Utilities' and Commissions'
Opinions on AFUDC Theory . . . ... .181

Summary . . . . . . . .... . 183









Page


CHAPTER VII SUMMARY AND CONCLUSION . . . .. .184

APPENDIX CHAPTER IV . . . . . . . .. 188

APPENDIX CHAPTER VI . . . . . . . ... 196

BIBLIOGRAPHY . . . . . . . . . . 220

BIOGRAPHICAL SKETCH . . . . . . . .. .224
















LIST OF TABLES


Table Page

1. RETURN ON CWIP WHEN CWIP IS INCLUDED IN THE RATE
BASE WITH AFUDC CREDIT TO REVENUE REQUIRE-
MENTS . . . . . . . . . . 48

2. RETURN ON PRIOR AFUDC (P) WHEN CWIP IS INCLUDED
IN THE RATE BASE WITH AFUDC CREDITS TO REVENUE
REQUIREMENTS . . . . . . . ... .50

3. RATES ASSUMED FOR SIMULATION . . . . . 59
RB RATE BASE
4. CAP RATE BASE RATIO OVER SIMULATION PERIOD . 64

5. RATIO OF CAPITAL COSTS AND RATIO OF REVENUE
REQUIREMENTS FOR ALTERNATIVE CWIP METHODS . 67
RB OPERATING INCOME
6. CAP OPERATING INCOME RATIO FOR ALTERNATIVE CWIP
CAP OPERATING INCOME. . . . . 71
METHODS ......... 71

RBNI
7. n- RATIO FOR ALTERNATIVE CWIP METHODS . . 72
CAPNI
NI
RB CASH FLOWS
8. CASH FLOWS RATIO OF ALTERNATIVE CWIP
METHODS ........... 75

9. TIMES INTEREST EARNED RATIOS FOR THE ALTERNATIVE
CWIP TREATMENTS . . . . . . ... .81

10. EFFECT ON RETURN ON EQUITY OF NOT COMPOUNDING
PRIOR AFUDC . . . . . . . . . 83

11. DIVIDENDS AS A PERCENTAGE OF NET INCOME EXCLUD-
ING AFUDC AND AFUDC AS A PERCENTAGE OF NET
INCOME CAP FIRM . . . . . . . .. 85

12. VARIABLE DEFINITIONS . . . . . . . 96

13. PARAMETER ESTIMATES FOR MULTIPLE REGRESSION
MODELS 100 ELECTRIC UTILITIES, 1972 . . .. .99

14. CROSS-SECTIONAL ESTIMATES OF THE PARAMETERS FOR
1970-1977, 100 ELECTRIC UTILITIES . . .. 100









Table Page

15. CORRELATION MATRICES FOR MODEL: P/B = a0 +
alBKYLD + a2GROBR + a3AFD + a4FT
100 ELECTRIC UTILITIES 1970 1977 . . .. 101

16. MEAN VALUE OF AFUDC/NI RATIO AND REGRESSION
COEFFICIENT FOR AFD, 100 PUBLIC UTILITIES,
1970-1977 . . . . . . . ... 103

17. THE EFFECT OF AFUDC ON THE COST OF CAPITAL,
100 ELECTRIC UTILITIES . . . . ... .105

18. PARAMETER ESTIMATES FOR VARYING PERCENTAGES OF
AFUDC/NI: 100 ELECTRIC UTILITIES, 1972 . . 107

19. CHOW TEST FOR EQUALITY OF SUBSET COEFFICIENTS 111

20. CROSS-SECTIONAL ESTIMATES OF PARAMETERS FOR
1970-1977, 100 ELECTRIC UTILITIES . . .. .112

21. THE EFFECT OF AFUDC ON THE COST OF CAPITAL,
100 ELECTRIC UTILITIES . . . . ... .114

22. MEAN P/E RATIOS, MEAN AFUDC/NI, T-STATISTICS
AND SIGNIFICANCE LEVELS FOR TWO GROUPS OF
FOURTEEN FLOW-THROUGH ELECTIRC UTILITIES.
GROUPS REPRESENT THE HIGHEST 30% AND LOWEST
30% OF COMPUSTAT'S FLOW-THROUGH ELECTRIC
UTILITIES RANKED ON AFUDC/NI. 1960 1975 .118

23. MEAN P/E RATIOS, MEAN AFUDC/NI, T-STATISTICS
AND SIGNIFICANCE LEVELS FOR TWO GROUPS OF
TWENTY-ONE NORMALIZING ELECTRIC UTILITIES.
GROUPS REPRESENT THE HIGHEST 30% AND LOWEST
30% OF COMPUSTAT'S NORMALIZING ELECTRIC
UTILITIES RANKED ON AFUDC/NI. 1960 1975 119

24. PARTIAL INCOME STATEMENT FOR ELECTRIC UTILITIES
UNDER THE JURISDICTION OF THE FEDERAL ENERGY
REGULATION COMMISSION SHOWING TREATMENT OF
CONSTRUCTION FUNDS . . . .. ..... .136

25. EFFECTS OF DIFFERENT TREATMENTS OF CONSTRUCTION
INTEREST TAX BENEFITS ON NET OPERATING INCOME
AND RETURN ON RATE BASE . . . . .. 141

26. CWIP ACCOUNTING METHODS, SIXTY-THREE
ELECTRIC UTILITIES, 1977 . . . . .. 150


viii









Table Page

27. THE DIFFERENCE IN AFUDC RATE AND THE ALLOWED
RATE OF RETURN ON THE RATE BASE FOR EIGHTEEN
FIRMS WHICH INCLUDE CWIP IN THE RATE BASE
WITH OFFSETTING AFUDC CREDITS TO REVENUE
REQUIREMENTS AND COMBINATIONS WITH THIS
METHOD . . . . . . . . . 151

28. PERCENTAGE OF CWIP IN THE RATE BASE FOR TEN
FIRMS WHICH USE METHOD 1 IN COMBINATION WITH
METHOD 3 . . . . . . . . .. 153

29. METHODS USED TO DETERMINE AMOUNT OF CWIP IN THE
RATE BASE . . . . . . . . .. 160

30. TREATMENT OF INTEREST TAX BENEFITS DUE TO
CONSTRUCTION WORK IN PROGRESS OF SIXTY-THREE
ELECTRIC UTILITIES, 1977 . . . . .. 163

31. TREATMENT OF INTEREST TAX BENEFITS DUE TO CWIP
BY SIXTY-THREE ELECTRIC UTILITIES CLASSIFIED
ACCORDING TO CWIP ACCOUNTING METHOD, 1977 164

32. NUMBER OF FIRMS WHICH DO NOT COMPOUND PRIOR
AFUDC . . . . . . . . . .. 168

33. STATE COMMISSIONS WHICH DO NOT PERMIT COMPOUND-
ING OF PRIOR AFUDC, CLASSIFIED ACCORDING TO
ALLOWED ACCOUNTING METHOD . . . . .. .170

34. CWIP AS A PERCENTAGE OF TOTAL ASSETS . . .. .175

35. REGULATORY POLICY VS UTILITIES' PRACTICE . . 177

36. COMPARISON OF UTILITIES' AND COMMISSIONS'
OPINIONS ON AFUDC THEORY. SIXTY-FOUR ELECTRIC
UTILITIES AND THIRTY-FOUR PUBLIC SERVICE
COMMISSIONS, 1978 . . . . . . . 182

37. TIME DIAGRAM FOR CAPACITY AND DEMAND FOR THE
FIRST TEN YEARS OF THE SIMULATION LEAD-IN-
PERIOD . . . . . . . . ... . 192

38. CAPACITY INVESTMENT SCHEDULE FOR SIMULATION
LEAD-IN-PERIOD . . . . . . . .. 193

39. BALANCE SHEET GENERATED BY IFPS . . . .. .194

40. INCOME STATEMENTS GENERATED BY IFPS. . . .. .195









Table Page

41. COMPARISON OF RESPONDING AND NONRESPONDING
ELECTRIC UTILITIES . . . . . . .. .197

42. RESPONDING FIRMS . . . . . . ... .198

43. QUESTIONNAIRE ON ACCOUNTING FOR CONSTRUCTION
WORK IN PROGRESS . . . . . . ... 200

44. COMPARISON OF RESPONDING AND NONRESPONDING
PUBLIC SERVICE COMMISSIONS . . . . .. .213

45. RESPONDING PUBLIC SERVICE COMMISSIONS ... .214

46. PSC QUESTIONNAIRE. . . . . . .. .215

47. FACTORS TO CONSIDER IN DETERMINING THE AMOUNT
OF CWIP TO INCLUDE IN THE RATE BASE ... .219
















LIST OF FIGURES


Figureage

1 DIVIDENDS AS A PERCENTAGE OF NET INCOME
AVAILABLE FOR COMMON, EXCLUDING AFUDC, FOR
THE ELECTRIC UTILITY INDUSTRY 1960-1977 . . 8

2 RATE MAKING FORMULA DERIVED FROM UTILITY'S
FINANCIAL STATEMENTS . . . . . . 41

3 THE EFFECT OF AFUDC ON TIMES-INTEREST-EARNED
RATIO: THE RETURN ON EQUITY CAPITAL EARNED AND
THE RESULTANT TIMES INTEREST EARNED RATIO WITH
VARYING PERCENTAGES OF AFUDC EXCLUDED FOR RATIO
CALCULATIONS . . . . . . . . 54

4 MACRO-LOGIC OF CONSTRUCTION PROGRAM MODEL OF AN
ELECTRIC UTILITY . . . . . . .. .57

5 RATE BASES OF ALTERNATIVE CWIP METHODS ... .62

6 UTILITY RATES FOR ALTERNATIVE METHODS OF CWIP
ACCOUNTING . . . . . . . . .. 65

7 OPERATING INCOME AND NET INCOME OF ALTERNATIVE
METHODS OF CWIP TREATMENT . . . . .. 69

8 CASH FLOW AFTER DIVIDENDS ....... . 74

9 INTERNALLY GENERATED FUNDS AS A PERCENTAGE OF
ANNUAL INVESTMENT IN CWIP FOR ALTERNATIVE CWIP
TREATMENTS . . . . . . . . .. 77

10 ANNUAL INVESTMENT IN CWIP AND CASH FLOWS AFTER
DIVIDENDS FOR ALTERNATIVE CWIP METHODS . .. 79

11 TIMES INTEREST EARNED RATIO FOR ALTERNATIVE CWIP
METHODS . . . . . . ... . .. . 80

12 THE EFFECT OF REGULATORY LAG ON RETURN ON EQUITY:
THE RELATIONSHIP BETWEEN FIXED UTILITY RATES AND
FIXED RETURN ON RATE BASE . . . . .. 89









Figure Page

13 MEAN P/E RATIOS FOR TWO GROUPS OF FOURTEEN
FLOW-THROUGH ELECTRIC UTILITIES. GROUPS
REPRESENT THE HIGHEST 30% AND LOWEST 30%
FOR COMPUSTAT FLOW-THROUGH ELECTRIC UTILITIES.
1960 1975 . . . . . . . . . 120

14 MEAN P/E RATIOS FOR TWO GROUPS OF TWENTY-ONE
NORMALIZING ELECTRIC UTILITIES. GROUPS
REPRESENT THE HIGHEST 30% AND LOWEST 30%
RANKED ON AFUDC/NI FOR COMPUSTAT NORMALIZING
ELECTRIC UTILITIES. 1960 1975 . . .. .121

15 SIMPLIFIED BALANCE SHEET AND RELATIONSHIPS OF
ALLOWED RATE OF RETURN TO THE TOTAL COST OF
CAPITAL. .. . . . . . .. . 130

16 FEDERAL ENERGY REGULATORY COMMISSION FORMULA FOR
CALCULATION OF COMPONENT RATES OF ALLOWANCE FOR
FUNDS USED DURING CONSTRUCTION . . . .. .133

17 THE EFFECTS OF DIFFERENT VALUATION BASES ON
CUMULATIVE DEFERRED TAXES . . . . .. 147















Abstract of Dissertation Presented to the
Graduate Council of the University of Florida
in Partial Fulfillment of the Requirements for the
Degree of Doctor of Philosophy



ELECTRIC UTILITIES' ACCOUNTING FOR CONSTRUCTION
WORK IN PROGRESS: THE EFFECTS OF ALTERNATIVE
METHODS ON THE FINANCIAL STATEMENTS, UTILITY
RATES AND MARKET TO BOOK RATIO

By

Geraldine Westmoreland

December, 1979


Chairman: E. Daniel Smith

Major Department: Accounting


Electric utilities account for funds used during con-

struction usually by one of three methods or combinations

of these methods: 1. They capitalize an allowance for

funds used during construction (AFUDC) which is added to

the construction work in progress (CWIP) investment and

recognize an unrealized AFUDC income for a like amount.

This income is realized in cash over the useful life of the

asset by way of depreciation and rate base increment.

2. They include CWIP in the rate base and receive a current

cash return for construction funds. 3. They include CWIP

in the rate base but have an offsetting AFUDC credit to

revenue requirements. This study consists of three parts:









1. A computer simulation of two alternative methods of

accounting for CWIP. 2. Security market research on the

effects on the cost of capital. 3. Current industry

practices and regulatory policy surveys.

The two polar methods, AFUDC capitalization and rate

base inclusion are simulated under moderate growth assump-

tions employing a deterministic model. The lead-in period

of the simulation provides a realistic vintage of assets

with prior capitalized AFUDC. The 25-year simulation

compares identical firms which differ only in CWIP treatment.

The firm which includes CWIP in the rate base could be

compared with one which has switched methods from AFUDC

capitalization. The simulation shows the effects of the

CWIP accounting treatment on rate base, utility rates,

operating and net income, cash flows, internally generated

funds and interest coverage. The effects of AFUDC on

normalized accounting are also considered.

The hybrid method of including CWIP in the rate base

with offsetting AFUDC credits to revenue requirements is

modeled for the first time in the literature. The current

cash return and the AFUDC return with various assumptions

about the allowed rate of return and the AFUDC rate are

derived. The compounding of prior AFUDC when this method

is employed is fully explored.

The security market research consists of multiple

regression analysis and price/earnings trends. The effect

of CWIP accounting on the price/book ratio of 100 electric









utilities was ascertained employing a linear model. The

price/book effect was translated into effect on cost of

capital. In addition, an attempt was made to determine if

a threshold amount of AFUDC/earnings is necessary before

cost of capital is affected. The trend of price/earnings

ratios for two groups of electric utilities ranked on

AFUDC/earnings is tracked for 1960-1975.

Extensive surveys of the electric utility industry and

state public service commissions were run. Accounting

practices, methods used to determine the amount of CWIP

to include in the rate base, treatments of interest tax

benefits and prior AFUDC, and the AFUDC effect on financing

are covered. A comparison is made of industry practices

with commissions' stated policy.















CHAPTER I

ELECTRIC UTILITIES' ACCOUNTING FOR CONSTRUCTION WORK
IN PROGRESS: THE EFFECTS OF ALTERNATIVE METHODS
ON THE FINANCIAL STATEMENTS, UTILITY RATES
AND MARKET TO BOOK RATIO



Motivation for the Study

Edison Electric Institute, the principal association of

the nation's investor-owned electric utility companies,

following an intensive research project on the economic

growth in the future, came to the conclusion that revisions

in public policy will be needed to encourage capital form-

ation and private investment in the productive facilities

of electric utilities. The electric utility industry is

the most capital intensive industry in the United States.

It is heavily dependent on the capital markets and will be

competing with industries which are not regulated for capi-

tal funds. The financial health of industry is critical to

the economy and the nation's energy posture.

Accounting research can aid regulators in their policy

decisions. May and Sundem (30) in their discussion on re-

search for accounting policy make several points which are

relevant to policy research for regulated industries.

Utility commissions, like accounting policy makers, must

specify the set of acceptable accounting practices. Their









final choice depends on their collective choice rule and

social decision system. An important area of research which

can aid them in their choice is the development and refine-

ment of accounting alternatives. In developing their model

for accounting policy decisions, May and Sundem state:

A first step in predicting ultimate conse-
quences of policy decisions is to predict
their effect on financial statements. There-
fore, a prior research, which predicts the
financial statement effects of alternative
measurement and reporting rules, can be an
important research contribution. And where
possible, empirical research confirming or
describing actual financial statement ef-
fects is probably even more helpful. (30, 753)

They also point out that the focus of much of the

accounting research concentrates on only the users of fi-

nancial statement data with little emphasis of the effects

on non-users. In decisions concerning accounting methods,

the non-users, mainly the ratepayers, are an important con-

sideration. In particular, the effects of the policy on

rates is a relevant factor which can be tracked with finan-

cial statements and accounting data.

Regarding research strategies, May and Sundem propose:

A largely neglected (in recent times), but
potentially profitable, avenue of accounting
research is investigating the degree to
which actual accounting time series may be
expected to differ under various accounting
alternatives, given observable or even as-
sumed or simulated environmental conditions
under which firms operate. Although un-
appealing as a sole basis of choice among
accounting alternatives, such investigation
can make a potential important, though per-
haps prosaic, contribution. . .(30, 758)









This dissertation asserts that the same approach can be

used for research involving the effects of accounting policy

on public policy. Its intent is to investigate the effects

of the financial statements of an electric utility under two

accounting alternatives given a simulated environment and

to attempt to describe how the cost of capital has been

affected in the past by the accounting method commonly

employed.

A further motivation is the nature of the problem which

this study addresses. Until the late sixties and early

seventies the accounting method used by electric utilities

for their construction work-in-progress did not affect their

costs in a material manner. Economic conditions were such

that the allowed return was realized by most utilities. The

size and duration of their construction programs were such

as to not create undue financial hardships. As economic

conditions changed and allowed returns were not realized and

the size and duration of programs increased, the financial

integrity of the industry became threatened. Large expendi-

tures were made on nuclear plants which were under construc-

tion for eight to ten years. Most firms were not allowed

to earn a cash return on these funds due to the accounting

method required by regulatory commissions. Consequently,

the method employed to realize a return on construction funds

affected the cost of the funds.

In 1976, the Edison Electric Institute projected that

the investor owned electric utilities will need to make









construction expenditures of one hundred twenty-two billion

dollars through 1981. Of this amount, 40% will need to be

financed with internally generated funds, as compared to

approximately 33% which has been the average in recent

years. The ability to do this is to a large degree contin-

gent upon the regulatory treatment of the funds tied up in

construction programs. The construction programs have in-

creased in size and longevity, as well as in cost. The

Federal Power Commission's statistics of privately owned

utilities in the United States reports that during the per-

iod of 1960-1975 the nominal cost of newly installed gener-

ating capacity grew at a compound annual rate of 6.4%. In

regard to relative size, in 1960 construction was approx-

mately 5% of total invested capital for the industry. By

1975 this percentage had risen to around 20%. The results

of the 1978 questionnaire survey done in conjunction with

this study, utilities project that the percentage will

increase to 24% by 1987.



Statement of the Problem

The electric utility industry has a massive construc-

tion budget for the remainder of the century. Financing

this budget requires that internally generated funds be in-

creased and that externally raised funds be obtained at the

lowest possible cost. The accounting method used to recog-

nize a return on funds used during construction affects both

the internal generation of funds and the cost of capital.









Since the industry is a regulated monopoly, the interests of

the various consumer classes must be balanced with those of

the investors. Consequently, the effects of alternative

accounting methods on utility rates, as well as on financial

statements and the cost of capital, need to be predicted

and/or confirmed. The methods used in the past of capital-

izing an allowance for funds used during construction (AFUDC)

and recognizing a paper income for a concomitant amount has

been much discussed in rate cases and trade journals, but

in-depth analysis of the effects of the method over time is

lacking.

Since the early 1900's public utilities have been allow-

ed to recognize a return on funds which were employed in

their construction programs. Two polar methods to recognize

this return are (1) include CWIP in the rate base, or

(2) capitalize AFUDC.1 The rate base method includes the

construction work in progress in the rate base for purposes

of current revenue requirements. This results in current

customers paying for the cost of funds used in construction.

The capitalization method requires capitalizing the allow-

ance for funds used during construction (AFUDC) by adding to

the asset value of the construction work in progress a per-

cent of the construction funds. Income for the period is



A hybrid method is also widely used which includes CWIP in
the rate base, and subtracts AFUDC from revenue require-
ments. This method could result in the return being
divided between cash and AFUDC credits. This approach is
discussed in depth later.








credited for a like amount. The capitalization method

results in an income figure which is composed in part of

"operating" cash earnings and in part of "paper" earnings

which will commence to be realized in cash when the asset

goes "on line" and the greater depreciation expense and a

larger rate base return are included in the revenue require-

ments from the consumer. Proponents of this method claim

it results in future customers, who benefit from the con-

struction, paying for the cost of the funds.

The present criterion for inclusion of assets in the

rate base requires the assets to be "used and useful." To

a large extent commissions do not regard construction as

meeting this criterion and, consequently, the capitalization

method is widely employed. Results of an industry survey

done in connection with this research show that approximately

50 percent of the responding firms use the capitalization of

AFUDC exclusively.

Simulation studies (39) have shown that if the con-

sumer's time value of money is the same as the firm's, the

two methods would result in identical net present values.

If this is, in fact, the situation, the method of accounting

for funds used during construction is just a bookkeeping

method and of little economic consequence to the investor

or the consumer. The "paper" earnings are as valuable to

the investor as the operating earnings. The determination

of the consumer's discount rate is not, however, a feasible








approach for policy determination. In addition, there is

reason to believe that the investor discounts the quality of

the capitalized AFUDC earnings because:

1) the ultimate cash realization depends upon demand

during the constructed asset's life and timely

rate relief during this time. The uncertainty

attached to these two conditions makes it not un-

likely that the investor will perceive these

earnings as riskier.

2) Cash flow problems are associated with the capital-

ization method. All other things being equal, a

firm which employs the capitalization method must

finance externally more frequently than a firm

which has all cash operating earnings due to the

non-cash nature of AFUDC.

3) In recent years, dividends for many firms employ-

ing the capitalization method have been greater

than 100% of their cash income. If investors

perceive these dividends as riskier, they

will demand a higher return. Figure 1 shows

dividends as a percentage of net income available

for common excluding AFUDC for the privately

owned utilities during the period 1960-1977. In

1974 this percentage was 110% and was still over

100% in 1975.




8








ii0i
*'Q- ----------- - -- .''-------



-1 /






7C



Source: Federal Power Commission "Statistics of Privately
Owned Electric Utilities in the United States." (1)



FIGURE 1
DIVIDENDS AS A PERCENTAGE OF NET INCOME AVAILABLE
FOR COMMON, EXCLUDING AFUDC, FOR THE
ELECTRIC UTILITY INDUSTRY 1960-1977









4) Bond rating agencies disregard AFUDC in excess of

a certain percentage of income in calculating bond

interest coverage. Consequently, the allowance

method can result in a higher cost of debt.

The above mentioned ramifications of the construction

accounting method lead to the speculation that the method

by which the cost of construction funds is realized is not

neutral. It would not be unreasonable to speculate that

firms which have a large percentage of AFUDC earnings in

their income figure are viewed as more risky than firms

which have mostly operating income.

The following excerpt from an article which appeared in

the Wall Street Journal on February 23, 1977 (44), points

up another problem with the capitalization of AFUDC. Namely,

fluctuations in reported income occur frequently when new

plants go on-line, and even though operating revenues have

increased, reported net income may drop. The operating

revenues are not great enough to replace the high AFUDC

earnings. In the case below, financing plans had to be

postponed.



NORTHERN INDIANA PS DELAYS STOCK OFFERING
TO REPORT PROFIT DROP

Despite increased gas and electricity revenues,
January net income declined about $1 million
from January 1976, Dean H. Mitchell, chairman,
said. Per share earnings last month fell to
about 28 cents from 36 cents in January 1976.
In the 12 months ended January 31, share








earnings were $2.43 compared with $2.50
in the 12 months ended December 3, 1976,
he added.
Mr. Mitchell said the lower earnings
were caused chiefly by a reduction in the
utility's allowance for funds used to fi-
nance construction and start-up expense
and increased depreciation charges result-
ing from a new generating unit. The unit,
near Wheatfield, Indiana, went into opera-
tion in December. The increased operating
revenues were not sufficient in Northern
Indiana's case to offset the lost AFUDC in-
come and the increased depreciation expense
due to the new plant. (44, 37)

The nature and quality of the AFUDC earnings have been

the subject of legal action. In a class action suit,

Greenapple versus The Detroit Edison Company, Morgan Stanly

& Co., Inc., and Price Waterhouse & Co. (24) it was alleged

that the prospectus grossly exaggerated actual earnings and

the trend of operating earnings for the years 1967 through

June 30, 1972, due to their presentation which treated "pro-

jected future earnings as if they were actual earnings,

under the caption, 'Allowance for funds used during

construction.'" The case highlighted the fact that rate

making accounting has not resulted in disclosure accounting.

Detroit Edison, as had many utilities, built up a construc-

tion pyramid which could be deceptive to the novice investor.

This pyramid works as follows: When construction on a plant

is completed and goes "on line," the AFUDC earnings would,

in theory, be replaced with operating cash earnings.

However, if rate relief is not granted and demand is not

sufficient, earnings will fall. Frequently, as was the case








with Detroit Edison, the former AFUDC earnings were not re-

placed with operating earnings, but with new construction

AFUDC earnings, and these earnings were greater than the

plant could have generated in operation. When the construc-

tion pyramid slows down, the AFUDC earnings evaporate. The

actual trend in operating earnings has been "concealed,"

claim the plaintiffs in this case, by the increased new

construction put in progress each year. It was contended

that the magnitude of the AFUDC effect created a cumula-

tive income item which distorted the entire presentation

of actual earnings power of Detroit Edison.

The litigation described remains unresolved. The

complaint seeks damages of $77.5 million against return of

the shares sold (less any amounts received by the alleged

class or prior sale of their shares), together with interest,

attorneys' fees and other costs. The actions taken regard-

ing the case are as follows:

May 28, 1976, the Court denied defendants' motion
for summary judgment without prejudice to renewal
upon completion of discovery. On completion of
discovery, defendants renewed their motion for
summary judgment and plaintiff cross-moved for sum-
mary judgment.
September 23, 1977, the plaintiff moved for a
determination of whether the action should be main-
tained as a class action.
April 26, 1978, the plaintiff voluntarily agreed to
dismiss the action as against the Company's indepen-
dent accountants and the underwriter, subject to re-
instatement in the event of a trial or other
proceeding on the merits. (38, p. 10)

The rate base method has been gaining acceptance as

a means of relieving financial stress for the utilities.

Following extensive hearings, the Federal Power Commission










in late 1976 issued Order Number 555 which recognized that

justification does exist for rate base treatment for

specific items of CWIP such as pollution abatement and con-

version facilities. The reasoning was that the present

generation was imposing costs on future generations. This

order applies to all companies with FPC jurisdictional

assets. In individual proceedings, the Commission will

permit other CWIP in the rate base if the company is in

severe financial stress. The Federal Energy Regulation

Commission (FERC) which has superseded the FPC has contin-

ued these orders.

These problems point up the need for an investigation

into alternative methods of accounting for CWIP. Even

though the policymaker's decision model is not known, it is

not unlikely that the effects of the alternative methods on

financial statements, cost of capital and utility rates are

important inputs.

Concisely, the problem is as follows:

The allowance method used by electric
utilities to account for a return on CWIP
has resulted in material problems. Cash
flows and capital costs have been affected
to an unknown extent. One alternative
method, inclusion of CWIP in the rate
base, would eliminate these problems, but
it has been objected to on the grounds
that it is inequitable for current customers
to pay for funds which benefit only future
customers. There is a need to ascertain
the relationships of financial variables
involved in the alternative methods and
the effects these methods will have on
utility rates, capital cost and financial
statements.









The hypothesis of this research is that electric

utilities' accounting for CWIP has resulted in methods which

have embedded in them elements of business, financial, and

regulatory risk. Business risk, the uncertainty inherent

in projections of future operating income, depends upon

(1) the extent to which a firm builds fixed cost into its

operations and (2) sales volatility. Relatively small

changes in sales result in large changes in profits if

operating leverage is high. The electric utilities were

regarded for years as having little business risk, but a

combination of events in the 1960's and 1970's altered the

situation, produced sharp declines in their operating income

thereby increasing the industry's business risk (8). This

research shows that the accounting treatment of CWIP has

contributed to this increased risk. Financial risk, the

uncertainty inherent in the firm's ability to cover its

fixed financial charges and to provide a reasonable return

on equity, depends upon the extent to which a firm uses

fixed income securities--debt and preferred stock--to finance

its assets. Since AFUDC is not cash income, it is not avail-

able for interest and dividend payments. Bond indentures

and rating agencies usually exclude a substantial percentage

of AFUDC in calculating the times-interest-earned ratio.

Therefore, financial risk is affected by CWIP accounting.

Regulatory risk, which is inherent in regulation,

increases both the financial and business risk of utilities.

It is, in fact, the primary risk of the utility investor.









Regulatory risk is the risk inherent in regulation. The

firm's profits and investment decisions are subject to the

commission's control. Regulatory risk is present in all

regulated firms to varying degrees. Not being directly

observable or measurable, the best that can normally be done

to measure regulatory risk is to somehow approximate the

regulatory climate. This can be done by observing the past

history of rate cases considering such factors as level of

allowed return on equity, granting of interim rates, rate

base test period, time involved in reaching final decisions,

treatment of existing stockholders and certain allowed

accounting practices. Regulatory risk is characterized by:

(1) Regulatory Lag. This is the time required for the

company to ascertain statistics for the test year, pre-

pare and file the documents for increases and commission

delibration time.

(2) Attrition. This is the difference between the rate of

return on equity granted by the Commission and the

actual rate realized. For example, assume the realized

rate of return is equal to the allowed rate of return.

Assume with time, inflation combined with fixed utility

rates cause the actual realized return to decline. If

a rate increase is granted, the rate of return rises,

but if there is continuing inflation, the target is

not attained. Over time the average realized ROE will

be less than the cost of equity.









(3) Rate Level. There is a risk inherent in every rate

hearing that the allowed rate of return will be less

than the cost of capital. Also, there is a risk that

assets will be excluded from the rate base, or expenses

disallowed for revenue requirements.

(4) Expropriation of locked-in capital. Utilities must

frequently go to the securities markets due to their

obligation to serve and their dividend policy. If

a firm's stock continually sells below book value and

the firm is compelled to float security issues at a

discount, the locked-in equity holders suffer confisca-

tion of their equity. A confiscation risk also exists

to the extent that the commission can alter cost of

capital estimation techniques.

(5) Stock Price. The price of a regulated public utility

is held to book value in times of inflation, unlike

unregulated firms whose prices rise with inflation.

This is the result of commissions regulating in a

manner which the market to book ratio target is around

1.0.

Other studies (10) have shown that normalization

accounting has an effect on the cost of capital to a public

utility and thereby demonstrated that regulatory risk could

be reflected in an accounting variable. This research shows

that an electric utility's CWIP accounting policy is also

a risk variable. By showing how the accounting variables

and utility rates are interrelated using a deterministic








simulation model, it is evident that business and

financial risk are increased by use of the CWIP accounting

method. When regulatory risk is considered in conjunction

with the risk inherent in the methods themselves, it is

not surprising that the CWIP accounting variable is an

explanatory variable in past market to book ratios.

This research deals with revenue requirements which

are necessary to cover the cost of capital funds, not the

pricing of electricity, which is a separate issue. The

effect on utility rates dealt with in this research is that

of revenue requirements per unit of demand under alternative

accounting methods. Pricing issues, for example, marginal

and/or embedded cost, multipart and/or discrimatory pricing

are not the primary focus of this research.

The goals of this research are four fold:

(1) To explicitly state the theory underlying a return on

CWIP and their accounting methods.

(2) To describe the effects of the alternative methods on

financial variables and utility rates over time. By

defining the interrelationships of these financial vari-

ables it will be possible, using a deterministic model,

to determine whether the methods inherently contribute

to the business and financial risk of the utility.

(3) To describe the past effects of AFUDC on the market to

book ratio and to infer from this the effect of AFUDC

on the cost of capital to electric utilities.









(4) To describe the actual practices of the industry and

commissions in dealing with the problem of construction

funds.



Methodology

(1) The theoretical justification for allowing a return

on construction funds and the methods employed will be

explained using regulation, accounting, and finance

theory.

(2) A computer simulation model of alternative methods of

accounting for CWIP will track the relationships of

financial variables and utility rates over time.

(3) Multiple regression models will be used to determine

whether AFUDC/NI is an explanatory variable for the

market-to-book ratio. Since the market may not dis-

count small percentages, an attempt to find the

threshold will be made.

(4) P/E ratios will be examined for firms with high AFUDC/NI

ratios versus firms with low AFUDC/NI ratios. Statis-

tical test of significance will be run on the two

groups.

(5) Actual practices will be determined employing the

literature, FPC orders, and questionnaire surveys run

on publically traded utilities and the state public

service commissions.















CHAPTER II

PRIOR RESEARCH



Prior research relevant to this study can be classified

into three areas; accounting theory of capitalization, re-

search dealing specifically with electric utilities' CWIP

accounting, and cost of capital studies of public utilities.



Accounting Literature on Theory
of Capitalization

Accounting literature dealing with the theory of capi-

talization of funds used during construction spans the

period of the late 1800's to present day. Early literature

dealt primarily with the question of whether the means of

acquiring property should have any bearing on its value.

More recent works applicable to regulated utilities deal

with the income realization aspects. Guthrie (25), in

1886, a proponent of capitalization of "interest" on propri-

etor's capital during construction, argued that there were

two ways of acquiring property, by purchase or construction.

If the property were purchased the price would include

interest on all capital employed during construction. If

property were acquired through construction there would be

no objection to capitalizing interest on debt as part of








the cost. Why should one type of capital differ from

another? Dividends or any interest on proprietor's capital

employed should be as much a cost as if the property had

been purchased. The means of acquiring property should not,

in itself, have any bearing on its true value.

Opponents of capitalization were of the same opinion

as Guthrie in regard to value, but reached the opposite

conclusion on capitalization. Walton (45), in the

July 1916 Journal of Accountancy, reasoned that interest

was purely a cost of financiering, and it did not add

anything to the intrisic value of the plant. Many values

were possible for the same plant depending, not upon the

plant, but the financing method. He contended that interest

charges should be carried as a deferred charge, along with

other expenses of organizing, and amortized over five to

ten years.

D.R. Scott (41), in 1925, believed that custom ruled

against charging assets for income foregone by stockholders

during construction, but saw no objection to allowing the

long-term borrowing rate on loans being capitalized, pro-

vided the offsetting credit was to a reserve account. He

saw the concept of objective evidence as a guideline. If

a firm did its own construction, the rate of interest paid

should be competitively determined. Since the length of

the construction period was not so objectively determined,

the total amount might not be capitalized.









Paton (34) in ]920 published an article in the Journal

of Political Economy entitled "Interest as a Construction

Cost." He opposed the capitalization of interest during

construction for industrial enterprises, but was one of the

first to point out that it might be perfectly logical to

do so in the regulated public utility due to the regulatory

process. He objected to unregulated firms capitalizing

interest during construction on the basis of the entity

theory. If construction funds are secured by a corporation

through stock and bond issues, the question of how to

account for interest and dividends is to be answered by

viewing the enterprise as an operating unit and bondholders

and stockholders as capital investors who do not sell their

services to the corporation. If the corporation secures

funds from these investors, any disbursement during con-

struction would be a reduction of capital since there were

no earnings at this time. He contended that accounts

should not reflect an accrual for "interest" on investment

prior to operation since, from the viewpoint of the account-

ing unit, there was no real gain in any assets. This

interest represented an economic gain but it was not within

the accounting framework to record it. In ]952, Paton

altered his earlier opinions (35). He shifted his view

of the corporate entity towards a "residual equityholders"

position. Interest during construction was not a return to

a group of capital suppliers, but a service sold by the

creditors to the common stockholders. This interest was an









incurred cost to the advantage of the stockholder. A

reduction in corporate resources had not occurred because

of interest paid during construction. Paton was of the

opinion that the mode of financing a particular asset during

construction should be viewed as a general financing cost,

incurred to get the business operating, and charged to a

special account rather than to individual asset accounts.

This cost would be a permanent asset, similar to organiza-

tion costs, and not subject to amortization.

Bierman (5), in his 1965 Financial Accounting

Theory, argues for capitalization of interest during con-

struction including implicit interest on equity capital.

He contends that unless interest is capitalized on both

debt and equity during construction, assets and net income

of early periods will be understated and in later periods

overstated. Bierman suggests that an unrealized income

account be credited to stockholder's equity and the credits

be recognized in the income statement as the asset is depre-

ciated. This treatment would nullify the argument that

profits were created on construction.

The practice of crediting income for an imputed

amount when construction interest funds are capitalized is

in direct contradiction of the income realization concept

as expressed by the American Accounting Association Concepts

and Standards Research Study Committee in 1965 (4) and by

the APB Statement #4 in 1970 (3). However, the regu-

lated utility is in an unique situation. It must be able








to recover all costs of service from the consumer. It

could be argued that "rent" on capital funds, both debt

and equity, are legitimate costs and there is a reasonable

expectation that this cost can be recovered over the life

of the asset. However, the accounting literature reveals

that the nature and timing of this income for public

utilities, like unregulated firms, is also controversial.

Paton (35) views the transaction which increases the

cost of the asset through an imputed amount as giving rise

to additional investment which should be classified as

part of permanent investment, such as paid-in or capital

surplus. Walker (43) justifies the recognition as income

in the regulated situation on the analogy that the sale is

recorded as revenue at the point when collectibility is

reasonably assured. Since the time of capitalization is

the time commissions allow the imputed cost to be added to

the asset value, the collectibility of this imputed amount

is reasonably established. Bierman (5), as noted

earlier, contends the income should be deferred until the

asset is actually producing revenues.



Electric Utilities' Accounting for CWIP

The three most extensive works in this area are

Quick's dissertation, "A Computer Simulation Analysis of

Alternative Methods of Accounting for Utility Construction

Work in Progress" (39), Arthur Andersen's ]977 study for









the Federal Energy Administration (14) and Pomerantz

and Suelflow's book, Allowances for Funds Used During

Construction (37).



The Quick study. Quick (39) did an extensive computer

simulation of different accounting treatments for CWIP

used by electric utilities. His finding were:

1. The rate-payer's time value of money will affect
the preference for construction treatments. If
his/her discount rate is lower than the cost of
capital to the utility, then the present value of
future revenue requirements is lower for rate base
firms and greater for capitalizing firms.

2. Firms which capitalize AFUDC have cyclically fluc-
tuating utility rates.

3. Fluctuations caused by the capitalization of AFUDC
coincide with the directions of fluctuations
caused by the depreciation method.

4. Firms which capitalize AFUDC and employ deprecia-
tion methods other than normalizing rely more
heavily on external financing.

5. Early periods in the simulation result in different
relationships than later stabilized periods.

The research carried out for this dissertation extends

the work done by Quick. Assumptions regarding the switch-

ing of methods, compounding of prior AFUDC, rate base

patterns and deferred taxes on the debt component of AFUDC

are changed from Quick's assumptions to be in conformity

with current industry practices.



The Arthur Andersen study. The Federal Energy Admin-

istration commissioned Arthur Andersen & Co. to do a









background analysis of the effects of the inclusion of

construction work in progress in the rate base and the

normalization of all income tax costs of the electric

utility industry. Their analysis included a comparison of

the revenue requirements including CWIP in the rate base

with AFUDC capitalized as income. Both single property

addition and a continuing construction program with various

assumptions as to useful life, construction period and costs

were employed. The case which was somewhat similar to the

research carried out in this dissertation was the simulation

of revenue requirements for growing construction programs.

They did not employ a lead-in period for the simulation and

assumed different rates of return for the two methods.

When construction work-in-progress was included in the rate

base, the allowed rate of return was 8.9% as compared to 9.6%

for the capitalization method. While the cost of capital

will be different for firms using the different methods, the

comparison of methods is confounded by this difference in

rates of return. Conditions are not ceteris paribus.

They addressed the cost of capital question by looking

at the common stock prices following a regulatory change in

Missouri to exclude CWIP from the rate base. The two major

electric utilities were used as an illustration of inves-

tor's reaction. The decline in common stock price for both

companies relative to the Standard and Poor's Index was

observed immediately following the election and the week

after. Based on this, they estimated that the overall








cost of capital would be higher by approximately .30 to .70

percentage points if CWIP were excluded from the rate base

and AFUDC were capitalized.



Pomerantz and Suelflow study. (37) This study addresses

the theory underlying the capitalization of AFUDC and the

actual procedures for accounting for it. The conclusions

regarding the economic and accounting theoretical bases

were:

1. There is economic justification for AFUDC capital-
ization. Since the investor requires compensation
commensurate with the value of his current fore-
gone consumption, interest must be measured from
the time capital is committed to the construction
of productive facilities.

2. Even though AFUDC represents an imputed cost and
is a direct contradiction of the income realiza-
tion principle, the overwhelming importance of
including AFUDC in the public utility's total
asset cost base justifies its recognition.

Pomerantz and Suelflow describe the actual accounting

procedures in detail. They discuss the capitalization

rate, give guidelines to follow in seeking to establish a

rate and actually develop a procedure which is a weighted

moving average cost of capital.



Cost of Capital Studies Relevant to CWIP Accounting

Until recently, accounting and regulation literature

have dealt to a large extent with the questions of whether

the financing of an asset should affect its value, the nature

of the "income" or capitalization credit, and the internal









effects of AFUDC. In recent years the question has arisen

as to whether the AFUDC capitalization method itself has a

cost. The effects on the cost of equity and debt have

been raised by Gordon (23) and Pinches et al. (36).

Gordon (23) in his study on The Cost of Capital to

a Public Utility did extensive regressions to test theories

of capital markets with past market data (1958-1968) on

public utilities. Employing the discounted cash flow (DCF)

theory of stock valuation, he hypothesized that dividend

and earnings yield varies with the risk or uncertainty of

the dividend expectation. Among the risk variables that

may influence yield he included a quality of earnings index

which considered accounting variables which affect cash

flows, such as deferred taxes, flow-through accounting and

AFUDC. This variable was used in three models which he

employed without conclusive results in regard to signifi-

cance. In all of the models it was significant for some of

the years.

The effect of CWIP accounting on financial risk was

noted by Pinches et al. (36) in their study of fixed-charge

coverage as a determinant of electric utility bond ratings.

Multiple discriminant analysis was used to discriminate

and predict both Moody's and Standard & Poor's bond ratings.

Complete stepwise procedures were employed to select the

model. From nineteen variables, the following six were

chosen: regulatory climate, total assets, net income/total









assets, earnings before interest and taxes/fixed charges,

construction expenses/total assets, and 1970-1975 growth

rate in net earnings. AFUDC/net income was one of the

considered variables but was highly correlated with the

selected variables and not included in the discriminate

model. The effect on bond rating of the six variables was

as follows: The more favorable the regulatory climate, the

higher the bond rating. Except for Baa (BBB) group, the

larger the firm in total assets, the higher the rating.

The large average size for the Baa (BBB) group was caused

by the presence of several large firms including Con Ed of

New York and Detroit Edison. The higher rated firms tend

to be more profitable in terms of net income/total assets

and have higher fixed coverage levels (EBIT/fixed charges),

than lower rated firms. The higher rated firms tend to

have a higher rate of construction expenses to total assets.

This may be due to the fact that firms in the Aa(AA) group

tend to cluster in the Midwest and Southern regions of the

country--areas where the demand for electrical energy is

growing faster than the national average. Lower growth

rates in net earnings were experienced by the higher rated

firms. This seeming inconsistency was attributed to the

accounting treatment for two items. First, the heavier

capital expenditures experienced by higher-rated firms

indicate that relatively more generating capacity was being

placed into service by these firms. This would cause depre-

ciation expenses to be greater for higher-rated firms,









resulting in lower reported earnings and lower growth rates.

Second, an examination of the variable AFUDC/net income

indicated that AFUDC represented a larger percentage of

net earnings for lower-rated firms than for the higher-rated

firms. Pinches notes:

. in such situations, while total reported
earnings may be growing faster for lower-rated
firms, financial analysts rating electric utility
bonds recognized that the "quality" of earnings
growth was poorer since it was due to the
inclusion of large amounts of AFUDC (3, 47).

This discriminant model correctly predicts 70% of

Moody's ratings, 76% of Standard & Poor's, and 81% of the

ratings for those firms when both agencies assigned the same

rating.

In an attempt to obtain some insight into the relative

importance of the six variables a rank ordering according

to five criteria was done. The univariate F ratio, the

scaled weighted method, the forward stepwise and backward

stepwise methods, and the conditional deletion method were

the five criteria. The univariate F and stepwise forward

methods showed that fixed coverage was the most important

variable by itself for both Moody's and Standard & Poor's

bond ratings. However, in a multivariate framework when all

variables in the model are considered simultaneously (as

seen by the scaled weighted, conditional deletion, and

stepwise backward criteria), fixed coverage becomes one of

the least important variables and growth in net earnings

one of the most important variables. The reason fixed









coverage is less important in a multivariate context is

because of the intercorrelation between it and the other

five variables. Pinches claims the findings indicate that

attempts in electric utility regulatory proceedings to

specify exact fixed coverage ratios that must be achieved

in order to maintain (or secure) a given bond rating are

both short-sighted and incomplete.

Johnson (26) compared price/earnings and price/book

value ratios of two groups of companies which employed

different CWIP accounting methods and reported a signifi-

cant difference between the groups with the capitalizing

group having the lower ratios. Group one consisted of

45 companies which capitalized AFUDC and realized no cash

return on their CWIP. Group two consisted of firms which

included CWIP in the rate base with an offsetting credit

to revenue requirements for AFUDC. This method results in

the possible realization of some cash return on CWIP if

there is a difference in the AFUDC rate and the allowed

rate on the rate base and if the prior AFUDC accumulated

on CWIP is also allowed in the rate base. Johnson did not

note in his work how or if he could ascertain these two

conditions.















CHAPTER III


REGULATORY THEORY OF ELECTRIC UTILITIES'
CWIP ACCOUNTING PRACTICES



Introduction

In 1920, Paton (34) noted that the accounting for con-

struction work in progress for regulated industries is sub-

ject to a different logic than for unregulated firms due to

the nature of the rate making process. The literature is

scant on the theoretical development of construction account-

ing methods from regulatory theory. The objective of regu-

lation is to simulate competitive outcomes to achieve welfare

maximization. The actual practice which has evolved for

utility pricing is accounting oriented, based on embedded

costs rather than marginal cost. As noted in Chapter I, this

research does not deal with the pricing controversy, but is

concerned with revenue requirements which must be recovered

from the consumer regardless of the pricing scheme. The

Supreme Court established the philosophy of recovering all

historical costs from the consumer including a "just and

reasonable" rate for the cost of capital funds as a legal

principle underlying regulation with the Hope case. In

arriving at a "just and reasonable" rate for the capital









charges employing this embedded cost approach to pricing,

a situation unique to regulated industries is evoked

according to finance theory. The book value of the stock

will be equal to the market value if a firm consistently

earns a book rate of return equal to the cost of capital.

Thus, commissions have been supplied with a rule of thumb

for checking the accuracy of their estimate of the cost of

capital, namely the market to book ratio.

In this chapter, regulatory theory and the micro-

economic theory of marginal pricing are briefly addressed.

The embedded cost approach and the legal precedents for the

rate making formula are given. The sensitivity of utility

rates to changes in the rate base are derived under the

two polar methods of accounting for CWIP. A third method

commonly used in practice, which includes CWIP in the

rate base and subtracts AFUDC from revenue requirements,

is modeled with the effects of compounding and rate differ-

entials derived.



Regulatory Theory

The organization and management of public utilities

is, for the most part, private with the central economic

decisions subject to direct government regulation. The

four components which distinguish the public utility from

the other sectors of the economy are (28):

1. control of entry,









2. price fixing,

3. prescription of quality and conditions of service,

4. the imposition of an obligation to serve all
applicants under reasonable conditions.

From 1877-1934 the U.S. Supreme Court took the posi-

tion that there were certain more or less readily identi-

fiable industries "clothed" or "affected" with a public

interest "to justify" legislatures subjecting them to regu-

lation. In 1934, in Nebbia v. New York, (32) the Supreme

Court abandoned the historic distinctions. It held that

there was no longer any constitutional barrier to legisla-

tures imposing any types of economic regulation on any

industries within their jurisdictions, where in their judg-

ment, it would serve the public interest. Formerly, the

Fourteenth Amendment was used as protection from regulation

by industries other than the specifically identified "public

interest" industries. As a consequence of this case,

however, as far as the U.S. Constitution is concerned, there

is no longer any distinction between public utilities and

other industries.

Yet, there is such a thing as a public utility.

There remains a core of industries, privately owned and

operated in this country which, in principle, the primary

guarantor of acceptable performance is the regulatory

commission which controls entry, price, and conditions of









service. Kahn (28, 11) contends that economic justification

for inclusion in this category would involve the following:

1. These industries constitute a large part of the
infrastructure of economic development. They are
suppliers of essential inputs to other industries
and influence the size and growth of the entire
economy. Many are characterized by great econ-
omies of scale.

2. Many are classified as "natural monopolies."
Their costs will be lower if they consist of a
single supplier.

3. For many possible reasons, competition does not
work well.



Microeconomic Prescriptions

Microeconomic theory can provide regulators with

principles which are assumed to maximize consumer satis-

factions if followed. They are behavioral rules describing

how prices should be set and descriptions of the ideal

results that these rules are supposed to produce. However,

as Kahn notes:

. its principles alone do not provide a sufficient
set of policy rules for regulated industries. They
do not answer the question of how and by what institu-
tional arrangements those ideal results are to be
achieved (28, 18).

The essence of regulation is the replacement of com-

petition with governmental orders on price, quality and

conditions of service along with an obligation to serve.

The single most widely accepted rule for the governance

of the regulated industries is to regulate them in such

a way as to produce the same results as would be produced








by effective competition if it were feasible. If the

market is perfectly competitive, the central policy pre-

scription of microeconomics is to set price equal to short-

run marginal cost. Economic efficiency requires that price

equal to marginal instead of average total costs, as long as

demand has some elasticity so that the marginal benefit to

the consumer of the additional output (as evidenced in the

price they pay) is equal to the incremental costs of produc-

ing that output. The consumer will buy more (less) than the

optimum amount if price is lower (higher) than marginal cost.

In theory, it is short-run marginal cost to which price

should be equated because it is short-run marginal cost that

reflects the social opportunity cost of providing the addi-

tional unit that buyers are at any given time trying to

decide whether to buy.

Another important criterion of marginal cost is that

it should reflect causal responsibility. The cost should

include all and only costs imposed on the economy by the

provision of one additional unit. To the extent that main-

tenance, depreciation, cost of capital, and various other

overhead expenses are not a function of use, they do not

belong in the ideal short-run marginal cost. Depreciation

is to a large extent a provision of obsolescence, not

physical deterioration as a result of production. Regard-

ing capital cost, the minimum necessary return on capital

is a function of time instead of utilization. Therefore,

economically efficient pricing would include only the









depreciation which varies with use, based on incremental

future cost, and no capital return.

Short-run marginal cost pricing does not necessarily

mean that long-run or fixed costs will not be covered.

Depending upon the cost behavior of the industry and the

relation of market demand to plant or industry capacity,

fixed costs will be covered without ever explicitly being

included in price (22). If average total cost curve turns up

at any point of output, it will be caused by marginal costs

exceeding total unit cost. Beyond this point, marginal cost

pricing will more than cover average total cost for constant

or increasing cost industries.

Legal Precedents and Embedded Cost Approach

The actual procedures which have evolved in practice

are quite different from the economist's prescription (29).

The process of rate-making involves the determination of

revenue requirements for a test year which will cover

operating expenses, which include income taxes and depre-

ciation, plus a "just and reasonable" rate of return on

a "capital rate base."

The legal principles underlying the regulation of

public utilities rest primarily on two Supreme Court cases:

Bluefield Water Works & Investment Co. v. Public Service

Commission of the State of West Virginia (262 U.S. 679,

1923) (6), and Federal Power Commission v. Hope Natural Gas

Company (320 U.S. 391, 1944) (17).









The Bluefield case stated the standard against which

just and reasonable rates is measured

A public utility is entitled to such rates as will
permit it to earn a return on the value of the
property which it employs for the convenience of
the public equal to that generally being made at
the same time and in the same general part of
the country on investments in other business under-
takings which are attended by corresponding risks
and uncertainties. . The return should be
reasonably sufficient to assure confidence in the
financial soundness of the utility, and should be
adequate, under efficient and economical manage-
ment, to maintain and support its credit and enable
it to raise money necessary for the proper dis-
charge of its public duties. (Bluefield, 693)

The Hope case gives legal sanction to the covering

of all costs, including capital costs

From the investor or company point of view it
is important that there be enough revenue not
only for operating expenses but also for the
capital costs of the business. These include
service on the debt and dividends on the
stock. . By that standard the return to
the equity owner should be commensurate with
returns on investments in other enterprises
having corresponding risks. (Hope, 603)

Consequently, the rate-making process involves the

determination of a "just and reasonable" rate of return

applied against a "capital base." The determination of

both of these factors have been the subject of consider-

able controversy.

The "just and reasonable" rate of return to the equity

holder has been arrived at by two main approaches in regula-

tory proceedings in the past: "comparable earnings" and

"discounted cash flow." The "comparable earnings" approach

involves analyzing rates of return on the book value of









the equity capital of various industry groups, both regu-

lated and unregulated, and establishing a comparable

"investment risk." Significant problems were encountered

in this approach. Namely, the procedure leads to circular-

ity for comparison of rates of earnings with other

regulated companies and comparisons of rates of earnings

on regulated and unregulated companies are not valid due

to the difference in investment risk and accounting

practices.

Since the mid 1960's, the "discounted cash flow"

approach has been used as an alternative to the "comparable

earnings." The theory was first developed by John Buer

Williams (46) and later extended by Gordon. It contends

that investors discount the expected future cash flows

attributed to a security to arrive at its value. The cash

flows include dividends and eventual liquidation value.

The discount rate is a market determined rate that takes

into account current market conditions and investment

risks. This approach has resulted in the one rule of

thumb recommended by many experts based on finance theory.

If a company's allowed rate of return on book value is set

exactly equal to the investors' required rate of return

then the company's market value per share will approximate

its book value. Therefore, the price/book ratio of a

utility should be equal to 1.

Robichek (40) contends that maintaining market

value around book value is tantamount to converting an









equity security into a perpetual, subordinated bond and that

the capital asset pricing model would be a better approach

to measuring required return. The Public Utility Commission

in Oregon recently ruled that the capital asset pricing

model be the primary means of determining the required rate

of return on equity. The lack of accuracy in determining

beta, the subjectivity of the risk-free rate, the choice of

a market index, and the underlying assumptions of the model,

make this method controversial [Meyers (31) Breen and

Lerner (7)]. Oregon is an exception, not the rule, in

using this approach, even though it is being used increas-

ingly as substantiating evidence with the discounted cash

flow method in rate cases. Robichek noted that the problem

of providing a "just and reasonable" return on "plant under

construction" would still plague regulators with the capital

asset pricing model approach.

The determination of the "capital base" with the

embedded cost approach which is used in practice raises

questions as to what should be included or excluded from

the base, the test period to use, and the valuation basis.

In regard to what should be included or excluded, the

traditional criterion has been to include all assets which

are "used and useful." Yet, this description lends itself

to many interpretations. Construction work in progress

could be classified as "used and useful" if it could be

shown that the capitalization of AFUDC increases the cost

of capital. It could also be argued that CWIP is necessary









for the "going concern" concept and is vital for continu-

ing service for present customers. The test year approach

attempts to incorporate the future into the regulatory

scheme. The strict historical test year assumes the past

relationships among revenues, costs, and net investment

during the test year will continue into the future. To

implement it, a recent 12-month period is selected as the

test year and rates are designed so the "fair rate of

return" will be achieved. Due to accelerated inflation

since the late 1960's, utilities have increasingly been

allowed to use future test years. The valuation basis of

the rate base was extremely controversial prior to the

1944 Hope decision. Reproduction cost or "fair value"

had been employed in the past and endless contention exist-

ed over the proper valuation of sunk capital. In the

Hope decision, the Court asserted that it would no longer

insist on Commissions taking reproduction cost into account

in fixing rates. The result, not the method employed, was

to be controlling. As long as regulation treats the

investors sufficiently well, the Court will pose no parti-

cular base. Attention was then shifted from a preoccupa-

tion with the rate base to the question of "fair" rate of

return. The controversy over the rate base still exists,

however. With price levels increasing since the Hope

decision, some companies have been allowed to use "fair

value" or reproduction cost. (Of the 100 companies in









Solomon Brothers Electric Utility Common Stock Market Data,

thirteen use fair value rate bases.)

The calculation of revenue requirements which has

evolved in practice, and comprises the embedded cost approach,

can be illustrated by referring to a utility's balance sheet

and income statement. Figure 3 shows a firm with $500

total assets which have been financed equally by debt and

equity. Assuming that the "allowed" return is the

"realized" return, the method of calculating rates can be

seen by looking at the rate formula. This firm does not

include CWIP in its rate base. The rate base includes only

the net plant-in-service. The average revenue (P) can be

expressed as Equation 1:

average (p) revenue requirements (R) OE' + kaRB
revenue quantity demanded (Q) Q

(Equation 1)

where:

k = allowed rate of return on rate base. It is
a calculated as a weighted average of debt and
equity

RB = rate base assets which, in this calculation,
include only net plant-in-service

OE' = operating expenses which include depreciation
and income taxes.

Since depreciation and income taxes are related to

the problem addressed by this dissertation, they will often

be dealt with separately in the discussion of rate making.

















Balance Sheet (Beginning of Period)


Assets

Plant in service (net
of depreciation)

CWIP


Total assets



Income Statement (End of

Revenues


Operating expenses

Fuel
SOther operating expense
Maintenance
Depreciation
Income taxes
Other taxes

Total expenses

Operating income

Other income

Allowance for funds
used during
construction

Income before interest

Interest

Net income


$400


Liabilities and Equity

Debt (D) at 8% (kd) $250


100 Equity (E) estimated fair
S return of 12% (k) 2

Total liabilities
$50 and equity $S



Period) Rate-making Formula

$134 Revenue requirements (R)

= Demand (Q) X Rate (P)
= OE' + k RB

301 ($94) + ($40) = $134


s 8
12
10
24
10



40





101

50

20


= Operating expenses (OE')




Rate of return (ka) on rate
base (RB)


Rate of return (kAFUDC) on CWIP



Rate of Return on total assets:
(kaRB + kAFUDCCwIP)
kdD


S30

kaRB + kAFUDCCWIP = kaD + ke

(Equation 3)


FIGURE 2

RATE MAKING FORMULA DERIVED FROM UTILITY'S
FINANCIAL STATEMENTS








To more readily see their effects, the general formula will

be expressed as:

Revenue requirements (R) = operating expenses (OE) which are

exclusive of depreciation and income tax + depreciation (D)

+ federal income taxes (T) + allowed rate of return (ka) X

rate base (RB).

Equation 1 can be expressed as:

P = R = ka(RB) + D + T + OE (Equation 2)
Q Q
Referring to Figure 2, it can be seen that net cash

revenues have been $40 and $10 of AFUDC credits have been

earned this period. Equation 3 on Figure 2 gives the balanc-

ing equation.

The $10 AFUDC will be added to the CWIP value. When the

asset goes on-line and the CWIP account is closed to plant-

in-service, a cash return will begin to be realized through

depreciation expense and rate base return on this increment.

The point of contention which has arisen over construc-

tion work in progress is whether the CWIP should be allowed

in the rate base. If this is permitted, the revenue re-

quirements must net $10 for the return on CWIP. The current

rates would be adjusted upward accordingly. There.would be

an income tax effect since taxes would have to be paid on

this return. If a 50% tax rate is assumed, revenue require-

ments would need to increase $20 for a net $10 after-tax

return.









Effect of Rate Base on Utility Rates

It is apparent that the CWIP return will affect future

depreciation expense, income taxes, and capital return by

way of the rate base. The sensitivity of rates to a change

in rate base can be determined by the total differentiation

of rates with respect to the rate base.

Expressed in functional form

rates (P) = f[rate base (RB), operating costs (OE)
depreciation (D), taxes (T), and demand (Q)]

P = f[(RB), (OE), (D), (T), (Q)]

The total differential, which gives the change in utility

rates with respect to a change in the rate base can be

expressed as

(1) (2) (3)
dP f d(RB) + f d(OE) if do
d(RB) 3(RB) d(RB) + (OE) d(RB) + (D) d(RB)

(4) (5)

af dT af dQ
af dT + f dQ (Equation 4)
_T d(RB) + Q D(RB) (Euation 4)

Terms (2) and (5) are assumed to be zero.

af d(OE) 3f dQ
a(OE) d(RB) O aQ d(RB) 0

It is assumed that variable operating expenses (OE)

and demand (Q) are not affected by CWIP accounting changes

in the rate base. While demand does have some price elastic-

ity as witnessed by the decrease in consumption following

the oil embargo, it is considered to be slight. Consequent-

ly, for simplicity and lack of a good estimate, demand is

considered to be price inelastic. Employing these








assumptions, the difference in rates caused by a small

change in rate base can be expressed as:

dP af af dD af dT
d(RB) (RB) (D) d(RB) +T dRB) (Equation 5)
af
The first term on the right hand side (RB) measures

the direct effect of a rate base change on utility rates,

while the second two terms measure the indirect effects

of depreciation and taxes. By taking the partial of the

rate-making formula, the coefficients for Equation 4 can

be found.
ka (RB) + D + T + OE
Q

P ka DP 1 DP 1
3RB Q' 'aD =' and T

The sensitivity of utility rates to changes in the rate

base, assuming price inelasticity, can be expressed as

dP ka 1 dD 1 dT (E tion
+ + (Equation 6)
d(RB) Q Q d(RB) Q d(RB)
ka
The direct effect is then seen to be A dollar change

in the rate base would be modified by the quotient of the

allowed rate of return on the rate base divided by the

anticipated demand. The indirect effects would be the

differences in depreciation and taxes caused by a change in

the rate base, both modified by the reciprocal of demand.



Effect on Utility Rates Caused by
Rate Base Method

If CWIP is included in the rate base, the direct effect

of d(RB), is positive during the construction period, due








1 dT
to CWIP additions and the indirect effect on taxes Q d(RB)

will be positive due to larger capital costs. The differ-

ence in depreciaton, dD, will be zero. Therefore, the

difference in rates due to a change in the rate base for

any year in the construction period (c) can be expressed as:

dPt ka 1 dTt

d(RBt) Qt Qt d(RBt)
Over the life of the construction program the difference in

rates can be expressed as the intergral.
tr dP t 1 dT
t kt
dP t dt = dt + dt.
o Jd(RBt) o Qt o Qt d(RBt)

When the construction is completed, the sign of the

direct effect will change from positive to negative, since

CWIP additions to the rate base will cease, and depreciation

will begin on the asset. The difference in rates for any

year in the operating life can then be expressed as

dPt ka 1 dD 1 dT
S + + t
d(RBt) Qt Qt D(RBt) Qt d(RBt)
Over the life (n years) of the asset, the difference is

tdP t ka 1 d Dt
-- dt = -dt + ---- dt +
n d(RB) n Qt n Q d(RBt)

f 1 dTt

n Qt d(RBt)








Effect on Utility Rates Caused by
Capitalization of AFUDC with no
CWIP in the Rate Base

If CWIP is excluded from the rate base during the

construction period, there will be no effect on utility

rates due to a rate base change. However, when the con-

struction is completed, (t*) the rates will be affected as

shown in equation 5,

dPt* kat* 1 dDt* 1 dTt*

d(RB)t* Qt* Qt* d(RB)t* Qt* d(RB)t*

with all three terms being positive. Since the resulting

capitalized cost is greater under this method than with

the rate base method, due to the capitalized AFUDC, the

d(RB), dD and dT will all be larger with a greater fluctu-

ation in utility rates than the rate base method. In

order to make the tracking of these effects over time

feasible, a computer simulation model is developed for the

alternative methods.



Model of Hybrid Method: Inclusion of CWIP in
Rate Base with Offsetting AFUDC Credits

The capitalization of AFUDC on all funds used during

construction or the inclusion of all funds in the rate base

are the two polar methods. A combination of these methods

is frequently used in practice. In addition, a third

method is employed by many firms which can result in a

current cash return on CWIP even though AFUDC is capital-

ized. This method includes CWIP in the rate base, yet








deducts AFUDC from current revenue requirements. This

method can be explained employing the simple model developed

earlier for Figure 3. The accounting return which the

firm will receive on total assets each period when the rate

base (RB) excludes CWIP is: ROA = ka(RB) + kAFUDC(CWIP)

(Equation 3). The return on the rate base is ka(RB) and the

return on CWIP is kAFUDC(CWIP). Current revenue require-

ments (R) are operating expenses (OE') and capital return,

which is the return on the rate base (ka(RB)). If CWIP is

allowed in the rate base and AFUDC is subtracted from

revenue requirements, the new modified rate base is RB +

CWIP, and the AFUDC subtraction is kAFUD (CWIP). The capi-

tal revenue requirements (R') can then be expressed as:

R' = ka(RB + CWIP) kAFUDC(CWIP) (Equation 7)

Return on assets each period is the capital requirement (R')

and the capitalized AFUDC, or:

ROA = ka(RB + CWIP) kAFUDCWIP + kAFDCCWIP (Equation 8)

current cash return AFUDC return

which simplifies to:

ROA = k (RB + CWIP) (Equation 9)

This method results in a total accounting return on

assets equal to the allowed return on the rate base. The

form of this return, current cash revenues or AFUDC credits,

is a function of the allowed return on the rate base, ka,

and the AFUDC rate, kAUDC, as well as the CWIP base.

Table 1 shows the form of the return on CWIP when the rate















TABLE 1

RETURN ON CWIP WHEN CWIP IS INCLUDED IN THE RATE BASE
WITH AFUDC CREDIT TO REVENUE REQUIREMENTS



Equation for Return on Assets:

ROA = ka(RB + CWIP) kAFUDCCWIP' + kAFUDCWIP'

Cash return AFUDC Credits


Assumptions:


1. 2 asset firm: RB, CWIP

2. CWIP = CWIP', firm consistently treats
compounding of prior AFUDC.

3. Allowed return k is realized return.
a


Cash
k Return

a = AFUDC 0

a > AFUDC CWIP(ka kAFUDC)

a < AFUDC


AFUDC
Return

kAFUDCCWIP

kAFUDCCWIP

kAFUDCCWIP


Total Accounting
Return on CWIP

kaCWIP

kaCWIP

k CWIP
a









differs, assuming the CWIP base is the same for the rate

base as for revenue requirement reduction.

If the CWIP amount in the rate base includes prior

period's AFUDC, the firm is compounding AFUDC regardless

of whether the AFUDC subtraction from current revenue

requirements, kAFUDCCWIP, contains prior AFUDC or not

Consider Equation 10:

ROA = ka(RB + CWIP) kAFUDCCWIP' + kAFDCCWIP'

current cash return AFUDC return

Note that this equation is different from Equation 8

in that the CWIP base for current revenue requirement re-

duction is CWIP.' and the rate base CWIP is lacking the

apostrophe. These bases may be different. The term

(RB + CWIP) is the rate base and it was shown in Equa-

tion 9 that the return on total assets was k (RB + CWIP).
a
If the CWIP amount in the rate base contains prior period's

AFUDC equal to P, then the overall compound element received

will be equal to k P. If the subtraction from revenue re-
a
quirements kAFUDCCWIP' does not contain P in the CWIP base,

then the compound element kaP will be received as a cash

return in this period's current revenue. The cash return

due to compounding of prior AFUDC is also a function of

difference between the allowed return on the rate base (ka)

and the AFUDC rate (kAFUDC). The possible forms of return

on prior period's AFUDC are shown in Table 2.

It can be seen that if prior AFUDC is consistently

included in the rate base and the revenue requirement, the















TABLE 2

RETURN ON PRIOR AFUDC (P) WHEN CWIP IS INCLUDED
IN THE RATE BASE WITH AFUDC CREDITS
TO REVENUE REQUIREMENTS


Equation for Return on Assets:

ROA = k (RB + CWIP) kAFUDCCWIP' + kAFUDCCWIP

Rate Base

cash return AFUDC Credits

Assumptions: L. 2 asset firm: RB, CWIP
2. allowed return realized return
3. P = prior AFUDC

Total Return on prj
Accounting (compound inc
Return Cash
CWIP CWIP' k on P Return

+P +P a=AFUDC k P 0
a

+P +P a>AFUDC kaP P(ka kAFUC)

+P +P a

+P -P a=AFUDC k P k P
a a

+P -P a>AFUDC k P k P
a a
+P -P a a a
-P +P a=AFUDC 0

-P +P a>AFUDC 0

-P +P a


ior AFUDC
crement)
AFUDC
Credit

kAFUDCP

k P
AFUDC

k AFUDC
AFUDC


0

0

kAFUDC

kAFUDCP

kAFUDCP








form of the return depends upon the rates ka and kAFUDC'

However, if the prior AFUDC is inconsistently included in

either of the CWIP amounts, the returns will vary. If P

is ever included in the rate base CWIP, the total accounting

return will always be k P with the amount and form varying

according to kAFUDC and CWIP', the revenue requirement

amount. If P is excluded from the rate base CWIP, prior

AFUDC is never effectively compounded.



Effect of AFUDC on Interest Coverage Ratio

One indicator of financial risk used by bond rating

agencies is a firm's interest coverage ratio. While the

exact weight attached to this ratio is not known, it cannot

be denied that the ability to service debt will affect its

cost. Since AFUDC is not a cash earning, it is not avail-

able to pay interest charges regardless of a frequent

practice of including a small percentage for coverage

purposes.

Using the embedded cost approach to utility pricing

and assuming a typical capital structure, the sensitivity

of coverage ratio to rates of return and AFUDC percentage

can be derived. This sensitivity is evidence that the

CWIP accounting method can affect the financial risk of a

utility.









Assuming the following:

TIE = time interest earned

EBIT = earnings before interest and taxes

kd = cost of debt

D = % debt in capital structure

k = rate of return on common equity

E = % earnings equity in capital structure

t = tax rate

A = AFUDC

the effect of AFUDC on the time interest earned ratio can

be derived as:

EBIT = kdD + k e + taxes

taxes = (EBIT A)t kdD t

Let EBIT A = EBIT*
EBIT*
TIE = -k
kdD

kdD + keE + (EBIT A)t kdD t A
TIE =
kdD

EBIT* keE (EBIT)t kdD t A At
kTD- kdD kdD

k E (EBIT)t kdD t A(] + t)
TIE ] kdD kdD kdD (Equation ]])


The effect of AFUDC on the coverage ratio is
therefore A(t + ])
therefore The tax term reflects the fact that

taxes are not included for AFUDC in revenue requirements

and therefore cannot be added back for coverage purposes.

The sensitivity of TIE to rates of return on equity

and AFUDC can be shown by assuming a capital structure








and embedded cost for debt and preferred stock. Figure 4

shows a firm with 55% debt and 45% equity which is composed

of 30% preferred and 70% common. Given the embedded cost

for debt and preferred, and assuming different rates of

return on common equity from 7% to 16%, the TIE ratio was

calculated for different percentages of AFUDC. As can be

seen, if a 14% return on common is allowed and anticipated,

close to 42% of EBIT can be AFUDC before coverage falls

below 2 times. However, if the allowed 14% rate is not

realized, but instead a 9% rate is realized, coverage falls

to around 1.7 times if AFUDC stays the same percentage of

earnings. However, with this situation, AFUDC as a

percentage of earnings would rise, making the coverage

even less.



Market to Book Ratio as a
Cost of Capital Indicator

Employing a situation unique to regulated utilities,

namely that profits are based on book values, and using the

discounted cash flow theory for stock price valuation, it

can be shown that the price of a utility's stock should

be equal to its book value if they have been allowed to

earn, and in fact, do earn their cost of equity capital.


















Times Interest
Earned Ratio

4X








3X
















Ix









10 20 30 40 50 60 70 80 90 10(

AFUDC as % of Earnings Before Interest and Taxes



FIGURE 3

THE EFFECT OF AFUDC ON TIMES-INTEREST-EARNED RATIO:
fIIE RETURN ON EQUITY CAPITAL EARNED AND THE
RESULTANT TIMES INTEREST EARNED RATIO WITH
VARYING PERCENTAGES OF AFUDC EXCLUDED FOR
RATIO CALCULATIONS









The relationship of the market price to book value is:

D1 (l-b)rB
P - g br (Equation 12)

where P = price per share

D1 = next period's dividends

k = required rate of return for equity

r = rate of return allowed by Commission

b = earnings retention rate

B = book value per share

if k = r, then P = B.

The relationship of market to book ratio when the allowed

return is not equal to the cost of equity can be expressed

as:

P/B ( b)r
k br

While this relationship is not linear, Brigham and

Bankston (9) have shown that for values of r within a

reasonable range around the appropriate value of k, that

P/B is linearly related to the allowed rate of return, r.

Gordon (23) has found risk factors in addition to return

on equity can be used to explain past P/B ratios. Conse-

quently, a hypothesis of this research is that the CWIP

accounting variable (expressed as AFUDC as a percentage of

earnings) has embedded in it elements of regulatory risk

which have affected the cost of capital and thereby the P/B

ratio.















CHAPTER IV

COMPUTER SIMULATION



Overview of the Model

The essence of the model is the period-by-period pro-

jection of key variables in simplified financial statements

based on assumptions about growth, costs, and accounting

methods. The Appendix provides a more detailed description

of the model and a sample of the financial statements gener-

ated. Figure 4 gives the macro-logic.



Model Inputs and General Assumptions

1. A specification of an initial balance sheet from a
95-year lead-in period. Typical of many utilities,
the model company has been capitalizing AFUDC during
its lifetime and normalizing income tax benefits due
to accelerated depreciation and the debt component of
AFUDC. A 60/40 debt to equity ratio is assumed. The
growth rate in demand during the lead-in period was 7%.

2. A specification of CWIP treatment. One of two methods
is used:
RB Method- includes the full amount of CWIP in the
rate base and earns the same rate of return on CWIP
as is earned on the rate base.
CAP Method- excludes CWIP from the rate base, capital-
izes an allowance computed as the product of CWIP and
the allowed AFUDC rate. This return is credited to
AFUDC income and added to the cost of CWIP.

3. A specification of the accounting procedures.

(a) The income tax benefits from accelerated depre-
ciation are normalized. Sum-of-the-years digits
is used for taxes and straight line is used for
rate purposes.












Initiate

Assumptions about financial
-. statement structure from
95 year lead-in period

Assumption about
-accounting methods

Assumptions about construction
-program, asset life
Specification of
growth rates in
demand and costs

Specification of
rates of return on
debt, equity, CWIP
and rate base

DO:

t= 1

to 25

Find capacity of generating plant to
be constructed. Find investment
each year (CWIP).

Find new financial statements Current
& update model parameters financial
+ statement
-values

End DO
I
Print End
Results



FIGURE 4


MACRO-LOGIC OF CONSTRUCTION PROGRAM MODEL
OF AN ELECTRIC UTILITY








(b) The interest tax benefits due to the construc-
tion program are normalized by the CAP method.
Taxes are deferred on the debt component of
AFUDC.

(c) The CAP method does not compound on prior AFUDC.
The annual charge for AFUDC is calculated by
multiplying the balance in the CWIP account
excluding any prior AFUDC by the AFUDC rate.

(d) The capitalized AFUDC is not deductible for tax
purposes in depreciation expense when the con-
structed plant goes on-line; however, it is
allowed for rate purposes. This results in a
different valuation base for tax and rate pur-
poses. This difference is amortized over the
life of the plant by adjusting income tax expense
and deferred taxes.

(e) Deferred taxes are subtracted from rate base
assets for determination of the final rate base
to be used for rate purposes.

4. A specification of the construction period and esti-
mated life of the generating plants. It is assumed
that it takes five years to construct a plant which
will last thirty years. The investment in construc-
tion is assumed to be equal in each of the five
years. One plant is always under construction.

5. A specification of anticipated growth in demand and
costs: Operating costs are assumed to be a specified
fraction of operating revenues. Construction costs
are assumed to be constant with the justification
that inflation is cancelled by economies of scale.

6. A specification of rates of return on debt, equity,
CWIP and rate base. Flexible rates are assumed so
that the allowed rate of return on the rate base is
achieved. (See Table 3).



Model Outputs

1. Balance Sheet, Income Statement to Stockholder's,
Income Statement to Internal Revenue Service.

2. Public Service Commission's rate base.





59









TABLE 3

RATES ASSUMED FOR SIMULATION



Allowed rate of return of equity: 13%

Allowed rate of return of debt: 9%

Allowed rate of return on rate base: 10.6%

Capital Structure 60/40 debt to equity

AFUDC rate: 10.6%

Tax rate:. .48

Growth rate in demand: 5.6%









Logic of Projection

The key idea is to start with a projected demand which

determines the size of investment. The capacity of each

new plant is such that five years after its completion the

combination of plants will have the capacity to generate

10% above the estimated demand at that time. The construc-

tion program is continuous. Each plant has a five year

construction period. As soon as one plant goes on-line,

another is started. Internally generated funds are the

first source of funds for the program. The remainder is

from external sources in accordance with the prespecified

capital structure of 60/40 debt/equity. Flotation costs

are assumed to be zero, new stock is sold at book value and

new debt is borrowed at the same rate as old.

The assumption of the accounting treatment of CWIP

will directly affect operating income, deferred taxes and

income taxes. The rate base and financial statements gen-

erated under the different accounting treatments will be

used for analyses of cash flows, utility rates and finan-

cial ratios.

The simulation assumes that two firms, RB and CAP,

start out in year 1, following a 95-year lead-in period,

with a group of identical assets with capitalized AFUDC.

For firm RB, the simulation is tantamount to a firm

switching CWIP accounting methods in year 1 following the

capitalization of AFUDC for 95 years.








The Interactive Financial Planning System (IFPS)

was used to simulate two utilities using different CWIP

accounting methods. The advantage of IFPS for the problem

addressed is that it has subroutines for depreciation and

amortization, and report generating capabilities which

lend themselves to the modeling of the different account-

ing methods. The investment strategy model and the

assumptions of the lead-in period are given in Appendix

along with sample financial statements.



Results of Computer Simulation



Effect on Rate Base

The rate base is responsible for many relationships

established by the simulation. Figure 5 shows that the

rate base of firm RB is continually rising in a smooth

pattern. This is due to the fact that the CWIP which is

added is greater than the annual depreciation. The CAP

firm has a fluctuating rate base. The fluctuations

increase in intensity over time. The decline in the rate

base pattern of the CAP method is due to the decreasing

net plant caused by depreciation. The sharp increases

are caused by plants going on-line and the accumulated

CWIP being added to the rate base. This sawtooth pattern

will be reflected in the utility rates, operating income,

and cash flows of CAP companies.






















Rate Base








340

RB



CAP
110




1 6 11 16 21 26 year






FIGURE 5

RATE BASES OF ALTERNATIVE CWIP METHODS









The relative sizes of the rate bases due to the alter-

native CWIP treatments over the simulation period can be

seen in Table 4. In the last three construction cycles

the RB firm has a lower rate base than the CAP firm during

the beginning of the cycle. This is caused by CAP's prior

AFUDC overwhelming the CWIP addition of the RB method.



Effect on Utility Rates

For the simulation period (25 years), five construc-

tion cycles of five years each were completed. Since the

lead-in period (95 years) assumed the firm was capitaliz-

ing AFUDC, the RB firm reflects a switch in method for

capitalization to rate base inclusion. (The fluctuations

of the CAP firm become symetrical above and below the RB

method following year 30 when the last of the lead-in per-

iod AFUDC effects are gone from the RB firm.)

Both methods result in fluctuating rates (Figure 6)

however, the fluctuations of the CAP firm are three times

greater than the RB firm. For the RB firm, the downward

drift in each cycle is caused by the increased demand.

This more than offsets the inclusion of CWIP in the rate

base. The fluctuation in utility rates for firm RB, when

a plant goes on line, is caused by the excess capacity.

The downward drift in the CAP firm is due not only to the

increase in demand, but also the decreasing rate base
















TABLE 4

RB RATE BASE RATIO OVER
CAP RATE BASE



YR 1 2 3
Ratio 1.00 1.15 1.23



YR 7 8 9
Ratio 1.08 1.16 1.26



YR 13 14 15
Ratio 1.11 1.22 1.34



YR 19 20 21
Ratio 1.19 1.30 .92



YR 25
Ratio 1.28


SIMULATION PERIOD


1.35



10
1.39



16
.92



22
.99


5
1.47



11
.95



17
1.00



23
1.07


6
1.00



12
1.03



18
1.09



24
1.16
















Utility Rates





3.0



2.0
\ \ CAP


1.0






1 6 11 16 21 26





FIGURE 6

UTILITY RATES FOR ALTERNATIVE METHODS
OF CWIP ACCOUNTING








which makes the revenue requirement for capital cost less.

(If demand growth were zero, rates would fall due to de-

creased capital cost.)

For the first two construction cycles (years 1 10)

the rates for firm RB are greater than (except for year 6

when they are equal) for firm CAP. Thereafter, in the

early years of the construction program, rates for firm CAP

are higher than firm RB. The relative size of the differ-

ences in revenue requirements due to the alternative methods

can be seen in Table 5. The greatest difference in

the two methods occurs in the last year of the first con-

struction cycle (year 5). The RB firm has a revenue

requirement for capital costs (return on rate base, depre-

ciation and income taxes) 1.47 times greater than the CAP

firm while the total revenue requirements are 1.16 times

greater.



Effects on Operating Income and Net Income

Figure 7 shows that the rate base pattern of the

alternative methods is reflected in the operating income.

The RB firm has an operating income which is smoothly

rising, as is net income. The CAP firm has a sharply fluc-

tuating operating income reflecting the rate base pattern.

The reported net income for the CAP method rises in steps

and is relatively smooth over the construction cycle even

though the rate base is declining. This is caused by the

AFUDC earnings which maintain the net income number even















TABLE 5

RATIO OF CAPITAL COSTS AND RATIO OF REVENUE
REQUIREMENTS FOR ALTERNATIVE CWIP METHODS



RB CAPITAL COSTS*
CAP CAPITAL COSTS


YR 1 2 3 4 5 6 7 8
Ratio 1.07 1.15 1.23 1.35 1.47 1.00 1.08 1.16



YR 9 *10 11 12 13 14 15 16
Ratio 1.26 1.39 0.95 1.03 1.11 1.22 1.34 0.92



YR 17 18 19 20 21 22 23 24
Ratio 1.00 1.09 1.19 1.30 0.92 0.99 1.07 1.16



YR 25
Ratio 1.28



*Capital costs include return on rate base, depreciation
and federal income taxes.















TABLE 5

(CONTINUED)



RB TOTAL REVENUE REQUIREMENTS R
CAP TOTAL REVENUE REQUIREMENTS



YR 1 2 3 4 5
Ratio 1.02 1.05 1.08 1.12 1.16



YR. 6 7 8 9 10
Ratio 1.00 1.03 1.05 1.09 1.13



YR 11 12 13 14 15
Ratio 0.98 1.01 1.04 1.03 1.11



YR 16 17 18 19 20
Ratio 0.97 1.00 1.03 1.06 1.10



YR 21 22 23 24 25
Ratio 0.97 0.99 1.02 1.06 1.10











H H
O S



i


1 H
0 2


\D L f l N r~









though operating income is dropping. The sharp fluctua-

tions noted in rates are reflected in the OI pattern of the

CAP firm.

Tables 6 and 7 show the ratios of operating income

and net income for the methods. The greatest difference

is noted in the first construction cycle (years 1 5).

The operating income in the last year of the cycle (year 5)

is 1.81 times greater for the RB method, while the net

income reflects only a difference of 1.18 times.

In the last construction cycle (year 21 25), the

ratio of the operating incomes has decreased with the CAP

firm having a greater operating income in the first two

years of the cycle (years 21, 22) and a reported net in-

come greater for the entire cycle. This reflects the cash

realization of prior capitalized AFUDC and a larger rate

base increment for the CAP firm.



Effect on Cash Flows after Dividends

The factors which cause the cash flow differences of

the two methods are:

1. Different returns on capital due to different

rate bases and, consequently, different utility

rates.

2. Different depreciation expenses due to the CAP

method containing capitalized AFUDC.
















TABLE 6

RB OPERATING INCOME
OPERATING INCOME RATIO FOR ALTERNATIVE CWIP METHODS
CAP OPERATING INCOME



YR 1 2 3 4 5 6 7 8
Ratio 1.12 1.25 1.41 1.59 1.81 .99 1.12 1.27



YR 9 10 11 12 13 14 15 16
Ratio 1.44 1.65 .91 1.04 1.19 1.36 1.56 .87



YR 17 18 19 20 21 22 23 24
Ratio .99 1.14 1.31 1.52 .86 .97 1.11 1.28



YR 25
Ratio 1.46
















TABLE 7

RNI
I RATIO FOR ALTERNATIVE CWIP METHODS
CAPNI



YR 1 2 3 4 5 6 7 8
Ratio 1.01 1.04 1.07 1.12 1.18 .90 .93 .97



YR 9 10 11 12 13 14 15 16
Ratio 1.02 1.08 .83 .86 .91 .96 1.02 .79



YR 17 18 19 20 21 22 23 24
Ratio .82 .87 .93 .99 .78 .81 .85 .91



YR 25
Ratio .98









3. Different income tax expenses due to the differ-

ences in net income, depreciation expense and

deferred taxes on the debt component of AFUDC.

Figure 8 shows the pattern of the cash flows for

the two methods. The RB method results in a cash flow

pattern which is a step function. The steps are caused by

the increased depreciation expense when a plant goes on-

line. Cash flows are greater for this method for all years

except for the first year of the last three construction

cycles (Table 8, years 11, 16, and 21). This reflects

the cash realization of prior AFUDC for the CAP firm.

The fluctuations in the CAP method cash flows are

caused by the changes in operating income due to the rate

base pattern. The peaks occur when plants go on-line and

greater capital allowances and depreciation expenses are

included in the rates. Since the model assumes actual

demand is equal to anticipated demand and instant rate

relief, the going on-line of a plant reflects an improve-

ment in cash flows for CAP firms. This is frequently not

the case. If demand is not sufficient and rate relief not

received when a plant goes on-line, the simulation results

are not substantiated.



Effect on Internally Generated Funds as a Percentage
of Annual Investment in CWIP

It has been projected (12) that in the next twenty

years electric utilities will need to generate 40% of their


















30

Cash flows
after
Dividends 25





20





15




10


\ I


I I
\


I~ 1



\ I


1 6 11 16 21 26


FIGURE 8

CASH FLOWS AFTER DIVIDENDS


CAP


















RB CASH FLOWS
CAP CASH FLOWS



YR 1
Ratio 1.09



YR 6
Ratio 1.03



YR 11
Ratio .99



YR 16
Ratio .96



YR 21
Ratio .95


TABLE 8

RATIO OF ALTERNATIVE



2 3
1.27 1.55


7
1.20



12
1.16



17
1.12



22
1.11


8
1.45



13
1.40



18
1.36



23
1.34


CWIP METHOD



4
1.96


9
1.85



14
1.78



19
1.73



24
1.71


IS




5
2.72



10
2.57



15
2.47



20
2.41



25
2.40








funds internally. Since 1970 the industry average has been

around 30%. The results of the simulation showed

an average consistent with the past industry average for

the CAP method, but not meeting the target percentage.

The RB method exceeds the target for all years.

The internally generated funds are derived from three

principal sources: (1) retained earnings, (2) depreciation

and amortization, and (3) deferrals of normalized items.

Since the CWIP treatment affects all three, the patterns

for percentage of funds generated internally for the al-

ternative methods are different over the individual con-

struction cycles and over the simulation period. Figure 9

shows the ratio of cash flows after dividends as a per-

centage of CWIP for the two methods. The RB firm (Fig-

ure 9A shows a constant ratio of internally generated

funds for each year of the five year construction period,

however, each cycle results in a slightly lower percentage.

The first five year construction cycle (year 1 5) had

63% of the funds generated internally, while the last

cycle (year 21 26) had 55% generated internally.

The CAP firm (Figure 9B) results in a declining

percentage over the five year cycle, but is constant for

each corresponding year in all cycles. When the annual

investment in CWIP includes AFUDC, the first year of the

construction program 53.3% of the funds are generated in-

ternally. The percentage decreases to 14.8% in the fifth










00:

N









-4 -4
<








HZ




-4
-4 >
iD -









I -o CN N E


*E- o<
Co c H z

ifl m N ZZ




0 cn
I- N



--r UU l


HZ~ 01: z
I__ U01 a0 0 J C
-~ C~E-o
0-I -C Z Z

41-O 0H
-44 W'' E- a,
44H u




a NC E-CZ'

U)
i l u? t l <


















On ( 4n N 1- z
C4 z













0000
ZLn

LnO
00U N












Ln H





00N
.1


L
-o C 1

'-0 04
* -------------- I us U < n M


















io
CD n


iD llC

on
H








year of the program. The average over the five year con-

struction program is 32.64%. Investment exclusive of AFUDC is

shown in Figure 9C.

The different patterns can be explained by Figure 10

which shows the annual investment in CWIP and cash flows

after dividends for the alternative methods. The CAP

firm (Figure 10B) has a smoothly increasing investment in

CWIP if AFUDC is considered, or a step pattern if AFUDC is

excluded. The cash flows after dividends are sharply fluc-

tuating as discussed earlier. This difference in cash

flows leads to a different percentage each year for the

ratio.

The RB firm (Figure 10A) has a step function for

both CWIP investment and cash flows. The CWIP pattern

reflects the simulation assumption of equal annual out-of-

pocket investment cost exclusive of the capital cost.

Consequently, an equal ratio is obtained for each year of

the cycle.



Effect on Times Interest Earned Ratio (TIE)

The times interest earned ratio is affected by the

CWIP treatment since part of AFUDC is generally excluded

from income available for coverage. There is an appre-

ciable difference between the range of the CAP method and

the RB method over the simulation period. The range of the

RB firm is between 2.61 3.28 while the CAP range is

1.36 2.62 (Figure 11). The fluctuations over the





















+
0
C4.
H H









N B mP


-4 aa
zo






C.)



r z
o 0a

r-





rl EQ
C\ 0









a Z Hm
Z Z0







_ IS = >r H
--- -i -I ^ -l








0D (fl ^r (N Hj


























3.0x









1. Ox





1 6 11 16 21 26




FIGURE 11

TIMES INTEREST EARNED RATIO FOR ALTERNATIVE
CWIP METHODS















TABLE 9

TIMES INTEREST EARNED RATIOS FOR THE
ALTERNATIVE CWIP TREATMENTS



YR 1 2 3 4 5
RB 2.89 2.83 2 2. .69 2.61
CAP 2.62 2.28 1.96 1.65 1.36
CAP* 2.64 2.36 2.01 1.71 1.51



YR 6 7 8 9 10
RB 3.06 1.97 2.88 2.80 2.70
CAP 2.62 2.28 1.96 1.65 1.36
CAP* 2.64 2.36 2.01 1.71 1.51



YR 11 12 13 14 15
RB 3.18 3.08 2.97 2.87 2.77
CAP 2.62 2.28 1.96 1.65 1.36
CAP* 2.64 2.36 2.01 1.71 1.51



YR 16 17 18 19 20
RB 3.27 3.15 3.03 2.92 2.81
CAP 2.62 2.28 1.96 1.65 1.36
CAP* 2.64 2.36 2.01 1.71 1.51



YR 21 22 23 24 25
RB 3.28 3.19 3.07 2.95 2.84
CAP 2.62 2.28 1.96 1.65 1.36
CAP* 2.64 2.36 2.01 1.71 1.51



*10% AFUDC/NI is included for coverage calculations.









construction cycles are also appreciably different. The

greatest fluctuation in the RB method for any cycle is

.46 TIE, while the CAP method results in fluctuations of

1.26 TIE. For the CAP firm, the TIE ratio is below 2.0

for three out of the five years in the construction cycle.

Table 9 shows a comparison of the ratios over the simula-

tion period. Since a small percentage of AFUDC/NI is

frequently allowed for coverage purposes, the TIE ratio for

a 10% inclusion is also given.



Effect of Not Compounding AFUDC

The model constraint is that the allowed rate of

return on the rate base is the realized rate of return.

Prior AFUDC is not compounded. Consequently, each period

some assets in addition to those financed by deferred taxes

are not receiving a return. This will affect the rate of

return on equity since the return on debt is fixed by con-

tract. The magnitude of this effect can be seen in Table 10.

The difference in the allowed and realized rates for each

year is the allowed rate of return on the rate base times

the allowance for funds used during construction

(ka X AFUDC). This can be readily seen by looking at the

theory of the model. Using the following definitions:

ka = allowed return on rate base which is a weighted
average of kd, the cost of debt, and kc, the cost of
equity.

kAFUDC = allowed return on CWIP

ka = kAFUDC





83









TABLE 10

EFFECT ON RETURN ON EQUITY OF NOT
COMPOUNDING PRIOR AFUDC



Years in
Construction Cycle 1 2 3 4 5

Return on Rate Base 10.6 10.6 10.6 10.6 10.6

Return on Equity
Allowed 13.00 13.00 13.00 13.00 13.00

Realized 12.72 12.18 11.51 10.70 9.80

Difference .28 .82 1.49 2.3 3.2









RB = Rate Base

TA = Total Assets

DT = Deferred Taxes

TA = RB + CWIP

RB = TA CWIP AFUDC DT(for the non-compounding
capitalizing method)

The total book return each period, in theory, is:

kaRB + kAFUDCCWIP = ka(TA DT)

Substituting for RB and recalling that ka = kAFUDC

ka(TA CWIP AFUDC DT) + kaCWIP = ka(TA DT)

kaAFUDC 0.

The difference between the return allowed on the rate

base and CWIP and the return allowed on assets exclusive of

deferred taxes will be kaAFUDC. Since the debt return is

fixed, this effect is reflected in the equity return.



Common Dividends Paid as a Percentage of Net Income
Available for Common Excluding AFUDC

Dividends as a percentage of net income excluding AFUDC

has increased dramatically for the utility industry since

1960. In 1974 and 1975, more than 100% of cash earnings

from operations was paid in dividends. The CAP firm shows

how this situation would be a likely occurrence. Table 11

shows dividends as a percentage of net income excluding

AFUDC and AFUDC as a percentage of net income. As can be

seen in the last year of the construction program, greater

than 100% of the net income is AFUDC.
















TABLE 11

DIVIDENDS AS A PERCENTAGE OF NET INCOME EXCLUDING AFUDC
AND AFUDC AS A PERCENTAGE OF NET INCOME
CAP FIRM


Year in
the Construction
Cycle

DIV.
NI AFUDC
AFUDC
NI


1 2 3 4 5


89% 120% 184% 405% undefined

21% 42% 62% 83% 105%