vicinity of Malden. The areas beyond Jackson and Malden are served by rural electric cooperatives and Missouri Utilities has not made any effort to duplicate distribution lines in those areas. Thus, it is highly unlikely that, in the foreseeable future, the Company will expand across the miles of the sparsely developed area near Jackson and Malden, add additional transmission lines, and loop them back into the system.

These transmission lines do not merely transmit power to Jackson and Malden, but are designed and constructed, at the specific request of these cities, to meet their needs within their respectively city limits. Malden's lines serve three distribution points, including the Malden generating plant. While the Jackson lines originally transmitted power to the Jackson power plant substation, additional lines have been added to meet the particular needs of the city and its residents. In 1965, at the city's request, Missouri Utilities built an additional line to serve a new industrial substation and relieve Jackson's own overloaded distribution circuits. In 1968, Jackson requested further extension of this radial transmission line to pick up the new West Substation, about a mile away. In short, the transmission lines in question, which benefit only the Cities of Jackson and Malden, have been designed and constructed at the sole request of the Cities to help these cities carry out their local distribution responsibilities.

This fact is expressly recognized in the contracts Missouri Utilities has executed with Jackson and Malden. The Cities have agreed to pay local facilities charges for the lines requested by them and built for their exclusive needs and benefits.

These contracts are, of course, still valid. During the remaining life of the contracts, they provide an additional reason for assigning the local facilities to the two cities. Because the contracts provide for the recovery of local facilities charges, the Company will completely recover its cost for them from Jackson and Malden. If the SFR-1 rate, as applied to Kennett, is determined with these lines rolled in, Kennett will be paying a rate designed to recover a portion of the costs of these lines. Thus, during the remaining period of the contracts, Missouri Utilities will over-recover the costs of the transmission lines in question.

The assignment of these local facilities to Malden and Jackson is approved.

F. Materials and Supplies

Cities' witness, Mr. Kumar, contended that the materials and supplies inventories included in rate base by the Company contain items that will eventually be used in construction of new plant rather than maintenance of existing plant in service. Following the theory that a utility is not entitled to include plant in its rate base until that plant is placed in service, Mr. Kumar computed an amount he considered equivalent to a normal 90 days of

materials and supplies for maintenance. He disallowed the rest as relating to construction and, therefore, not properly a part of rate base.

Missouri Utilities admits that some materials and supplies will inevitably end up in construction work in progress. However, the Company argues that if an item were to be used in CWIP, it would be replaced to maintain the inventories necessary for normal maintenance. Thus, unless the materials and supplies inventories are larger than necessary for maintenance functions, rate base will not be increased by supplies earmarked for construction. Furthermore, while CWIP is not includable in rate base, the utility may capitalize the return on investment in the form of an allowance of funds used during construction. AFUDC is calculated on jobs which exceed $10,000 and thirty days in progress. Materials utilized for these jobs are charged directly to the job and do not go through the material inventory. Thus, any significant amounts of materials and supplies carried into CWIP will not find their way into the materials and supplies inventories includable in rate base. Missouri Utilities' explanation is persuasive and the amount of materials and supplies is not excessive. The Company has established that the materials and supplies included in rate base are reasonable and proper.

IV

COST OF SERVICE A. Use of Division-Wide Costs

The Commission consolidated this docket with Docket No. ER77-348, concerning Missouri Utilities' rate filing for its Central Division, to determine whether the Company's cost of service should be determined for each operating division or on a company-wide basis. Although Missouri Utilities has reached a settlement with the only wholesale customer in its Central Division, the issue remains. Missouri Utilites has submitted cost of service data based on its costs for the Southeast Division. Staff supports the Company. Intervenor Cities contend that the cost of service should, instead, be based on Company-wide costs.

It is clear that the distinct nature of Missouri Utilities Central and Southeast Divisions merit isolation of Southeast Division costs in determining the cost of service here. There are no physical links between the two portions and no power is transmitted between them. The Company keeps separate books for each division. Customers are billed on different bases. Accordingly, use of Southeast Division costs in this proceeding is proper. B. Administrative and General Expense

Missouri Utilities has divided its administrative and general expenses into electric, other, and common categories. Those expenses in the common category are generally allocated between the electric and other utility departments on the basis of relative

revenues. Total electric administrative and general expenses are than functionalized by using the mean of three ratios-wholesale electric revenues, labor, and gross plant. Commission Staff contends that this method of allocating to functions is not based upon incurrence of these costs. It argues that there is no reasonable relationship between A & G expense and wholesale electric revenues and that, accordingly, revenue should not be used as an allocation factor. Staff asserts that the Company's use of gross plant as a basis for functionalization is also improper since most A & G expenses are labor intensive and not capital intensive.

Staff proposes a different method. It assigns each A & G expense item common to the electric and other departments by application of a labor ratio, with the exception of insurance which is assigned on a gross plant ratio. It then allocates a portion of each item of total electric expense to the various functions using a gross plant ratio for insurance and labor ratios for other categories. A few items of A & G expense are assigned entirely to the retail portion of Missouri Utilities' business. They are research, demonstration and development expense, regulatory commission expense, rate case expense, and interdepartmental rents.

Cities essentially agree with the Staff position. Whether or not it is appropriate to assign all regulatory commission expense to retail, they would eliminate this item from the Company's cost of service because Missouri Utilities' rate filing in this case is allegedly unwarranted.

In Missouri Power & Light Company, Opinion No. 31, Docket No. ER76-539, issued October 27, 1978, 5 FERC ¶ 61,086, the Commission affirmed the Administrative Law Judge's Initial Decision to the extent that it rejected the use of a method of allocation similar to Missouri Utilities' approach. The Commission noted that it is FERC policy to functionalize A & G expenses on the basis of either plant or labor ratios and agreed with the Staff position that functionalization on the basis of wage ratios is the most reasonable. This strong preference for labor ratios is also shown in Lockhart Power Company, Opinion No. 29, Docket Nos. E-9469 and ER76-377, issued September 22, 1978 4 FERC ¶ 61,337.

Missouri Utilities must establish that its method of allocating A & G expenses to functions is a reasonable one. The Company has clearly failed to meet this requirement. It does not establish by any representative factual showing that the items of A & G expense are closely related to revenues. Staff's approach, on the other hand, follows Commission policy and is reasonable under the circumstances. Staff's functionalization of A & G expense by use of plant ratio for insurance and labor ratios for other items is hereby adopted.

Staff contends that research, demonstration, and development expenses should be assigned entirely to retail customers. Most of Missouri Utilities'

expenses in this category consist of payments to the Electric Power Research Institute which calculates its assessments solely on the basis of retail sales. Staff's position, based upon Commission precedent,' is adopted.

Staff assigns regulatory commission expenses, rate case expense, and interdepartmental rents to the retail portion of Missouri utilities' business because the Company has failed to establish that these directly assignable expenses should be allocated. Missouri Utilities did not address this issue at the hearing or in its brief and, accordingly, has not rectified the deficiencies in its presentation. The Presiding Judge finds that these items of A & G expense must be eliminated from the Company's cost of service. Because of the disposition of this issue, it is not necessary to address Cities' treatment of regulatory commission expense.

C. Allocation of Transmission Expense

Missouri Utilities allocated transmission expense on the basis of kilowatt-hour sales. While the Company concedes that transmission expenses are generally considered to be demand related, it argues that the Cities' practice of generating at times of system peak and avoiding demand charges means they will not be charged to fair share of transmission expenses if the 12 CP demand method is used. Staff and Cities dispute Missouri Utilities' departure from the traditional approach.

Transmission expenses are demand related. Moreover, although Cities do attempt to avoid demand charges by generating their own power during heavy load periods, they imposed substantial demand on the Southeast system during the 12 CP of the test year. Thus, Missouri Utilities' fear that the wholesale customers will be attributed only negligibly costs despite almost continuous use in the system is unfounded. Utilization of the 12 CP method is ordered.

D. Interest Expense Synchronization

Staff witness Lehman adjusted interest expense to synchronize it with the current cost of capital. For short-term debt he used the average of 13 monthly balances of notes payable and placed the cost at 9%, the then-current prime rate. He calculated long-term debt by applying the weighted cost of long-term debt to the rate base, using updated figures to reflect the Commission's preference for recent data in determining capital structure. Cities support the Staff procedure.

While Missouri Utilities agrees that the theory of synchronization is a good one, it contends that Staff has overstated interest expense in this case. The Company claims that $6,000,000 in long-term Idebt which it issued toward the end of the test period was used to roll over and eliminate the shortterm debt as calculated over the course of the year.

It is normal for a company to use long-term debt to retire its short-term debt. The process is likely to be repeated frequently over time as short

term debt is used to finance CWIP, then rolled over as a project is completed. This fact makes Missouri Utilities' argument irrelevant. Although the particular short-term debt used in Staff's calculation may have been rolled over by long-term debt, other short-term debt will soon follow. The Company has not shown that this process is atypical in its case nor that the figures used by Staff are unusual. Accordingly, Staff's methodology is adopted.

E. Fuel Expense Allocation

Missouri Utilities allocated fuel expense, related to the Company's own generation, on a demand basis. Staff proposed that fuel expense be allocated on the basis of energy sales. At hearing, Missouri Utilities conceded the appropriateness of this adjustment.

Cities would eliminate the fuel expense entirely. They contend that none of it can properly be allocated to the wholesale customers because Missouri Utilities cannot generate unless directed to do so by Union Electric, its parent company. However, despite the claims of the Cities, it is clear that these Missouri Utilities' generating units are not controlled exclusively by Union Electric. As explained by Mr. Chapman, they were used during the test period to generate power for the Company's system. This generation entered the common transmission system where it was used by both wholesale and retail customers. The Presiding Judge finds that wholesale customers should be paying their proportionate share of this legitimate expense. Cities' proposal is rejected.

F. Synchronized Fuel Clause Revenues

Missouri Utilities' rate design includes a fuel adjustment clause which relies on the actual preceding month's fuel expense to compute the fuel adjustment component for the current month's billing. The result is that for the November 1975 to October 1976 test year, the Company has applied October 1975 to September 1976 fuel clause expenses. Staff witness Gekas synchronized the fuel adjustment revenues to match the test period. He recomputed the fuel cost component utilizing the fuel expense incurred and corresponding kWh sales during the test period. Intervenor Cities agree with Staff. Missouri Utilities did not object to the Staff proposal on brief. The adjustment is consistent with the test period concept and is adopted. G. Allocation of Payroll Taxes

Missouri Utilities functionalizes its test year social security and unemployment taxes on the basis of net plant. Staff utilizes labor ratios. Missouri Utilities does not oppose the adjustment in brief. Clearly, these taxes are more closely related to labor than to plant. The Presiding Judge adopts the Staff methodology.

H. Operating Revenues as Credit to Expense

The Commission Staff applied other operating revenues of the Company as credits to operating

and maintenance expenses. It contends that certain items of rental revenues, commissions and miscellaneous functional service charges should have been considered as electric operating revenues for purpose of the cost of service.

Missouri Utilities has not objected to the Staff adjustment. The Staff proposal is approved. V

ALLOCATION TO JURISDICTIONAL
CUSTOMERS

Missouri Utilities has allocated demand-related costs between jurisdictional and nonjurisdictional customers using the twelve coincident peak (12 CP) method. Staff supports this method to obtain consistency with the Union Electric allocation factors approved by the Commission in prior cases.

Cities also propose a demand allocator based on the 12-CP method. However, they contend that the amount of their net firm generating capacity should be deducted from the actual demand figures of the test period.' According to the Cities, this methodology is justified by the fact that they utilize their own generating capacity at times of Missouri Utilities' system peaks to avoid demand charges.* Cities argue that this practice of self-generation means that they are interruptible customers which impose no demand costs on the system. In the past they have removed themselves from the Missouri Utilities system at times when the Company was experiencing loads of at least 80% of annual peakthe times when demand charges were imposed. Cities contend that they can and will utilize their own generation at the 12 CP in the future to avoid demand charges, although they did not do so in the test period. They, therefore, assert that credit should be given for their own generating capacity in allocating demand costs.

The fact that the Cities have not been the cause of a demand charge component of Missouri Utilities' purchased power costs from Union Electric is not relevant to the allocation of fixed costs issue. These charges are assigned to the customers who cause them. They have, therefore, been assigned entirely to the retail customers.

However, the Cities' contention that they are interruptible customers is related to the appropriate allocation of demand-related costs, such as those incurred by Missouri Utilities in carrying out its transmission service. The Commission found in Minnesota Power & Light Company, Opinion No. 12, Docket No. E-8494, issued April 14, 1978, 3 FERC 61,045, at mimeo pp. 21-23, that certain partial requirements customers receiving supplemental power should not be allocated demand costs. Truly interruptible customers, who use only the capacity planned to meet the needs of firm customers and do not represent a firm obligation, need not pay for fixed costs.

The problem with the Cities argument is that they are not purely interruptible customers. The

SFR-1 rate which is the subject of this proceeding offers firm service to the Cities if they choose to take it. If they do take power at a time when Missouri Utilities' load is at least 80% of peak, they will pay certain demand charges. These demand charges represent the additional cost to the Cities for taking firm service. Yet, the rate does not require the Cities to take all their power from the Company. They may utilize their own generating capacity, as they have done in the past, to avoid demand charges.

The Presiding Judge finds that because the Cities are only interruptible customers at their own option, the general rule should be followed. The demand allocator should be based on the actual demand imposed on Missouri Utilities' system during the 12 CP of the test year. It is true that the Cities' practice of supplying their own needs at time of system peak affects the demand costs they impose on the system. However, accounting for this particular aspect of the service provided requires a flexibility that is better found in the design of the rate. Accordingly, the effect of this practice on the level of costs will be addressed at that stage.

VI

RATE OF RETURN

A. Cost of Long-Term Debt

Missouri Utilities and Intervenor Cities calculate the imbedded cost factor applicable to the Company's long-term debt to be 7.24%. Staff witness Hutt argues that the appropriate figure is 7.21% which he claims to be reflective of the capital structure existing on October 31, 1977. Mr. Hutt contends that his calculation is based on Missouri Utilities' response to Staff's Power Request No. 44, entitled "Missouri Utilities Company, Debt Capital." Mr. Hutt apparently infers from Power 44 that the Company has calculated the cost of debt based on net proceeds. He argues that it should be calculated using the amount of long-term debt outstanding, a position which Missouri Utilities readily concedes. However, when the correct method is used, as articulated by Mr. Hutt, one arrives at a cost of long-term debt of 7.24%. From the Company's Exhibit No. 2 it can be seen that Missouri Utilities has used the appropriate method for finding the cost of debt (using long-term debt outstanding) and has determined the correct figure. B. Return on Common Equity

Missouri Utilities requests a 16.09% return on common equity. In support of this proposal, the Company's expert witness, Dr. Michael D. Sherman, analyzed the performance of the common stock of Missouri Utilities' parent company, Union Electric.' He computed discounted cash flow (DCF) returns of 54 companies allegedly comparable to Union Electric. He then took the average of these returns as aplicable to Union Electric, which appears at the low end of the return spectrum,

and

made a further upward adjustment for flotation costs and "market pressure" to arrive at his recommended figure.

Dr. Sherman obtained his list of comparable companies by using a procedure called Stochastic Dominance. This method compares the amount of returns on common stocks to evaluate companies according to an investor's preference for increasing wealth and compares the variability of those returns to evaluate such companies in accordance with an investor's aversion to risk. Companies are then ordered in terms of their relative superiority for return and protection from risk. Both factors are measured. Those companies whose performance is clearly superior or inferior to Union Electric are eliminated. The "neutral" set that remains is comparable to Union Electric in that any difference in the risk of another company vis-a-vis Union Electric is accompanied by a corresponding difference in return. The comparable companies all satisfy a particular spectrum of risk-return judgment on the part of the investor. Dr. Sherman contends that this set of companies is appropriate for a comparative earnings analysis:

[T]he Hope and Bluefield cases define comparability in terms of the equivalence of the riskreturn relationship among business undertakings. Thus, comparability must be determined by a direct assessment of both risk and rate of return. It is neither risk alone nor return alone that establishes correspondence among firms; it is the combination of the two which delineates the set of equivalent enterprises. More specifically, it is the similarity of the adjustment process by which increased return compensates for increased risk exposure, and vice versa, that defines comparability. (Tr. p. 72)

Dr. Sherman then applied his own version of DCF analysis to the 54 comparable companies. The return he obtained for each company was the sum of (1) the average of the ratios of total annual dividends per share to year-end book value for the years 1967 through 1975 and (2) the compound annual growth rate of dividends per share for the period 1966 through 1975. Dr. Sherman argued that the use of a ratio of dividends to book was necessary to provide a rate of return which fairly compensates present investors in the company as well as those Missouri Utilities will seek to attract in the future.1 The average of the DCF returns for these companies was found to be 14.81%. An adjustment was made for flotation costs and "market pressure" to reach the recommended figure of 16.09%.

Because the companies in Dr. Sherman's comparable group are not similar solely on the basis of risk, the DCF returns covered a very wide range. The lowest was 2.5% (U.S. Steel) and the highest 30.4% (Lucky Stores). Missouri Utilities explains

that:

* what the DCF analysis shows is that U.S. Steel has a high probability of a low but stable growth rate as compared with Lucky Stores, Inc., which shows a low comparative probability of a high return, thus making their comparative probabilities of earning any given return or less equivalent. The DCF analysis with respect to these two stores [sic] confirms the viability of the SD analysis, demonstrating that the increased risk of Lucky Stores requires a high rate of return, while the decreased risk of a lesser return requires a lower DCF return. (CRB, p. 12)

This definition of comparability does not comport with the one envisioned by Hope and Bluefield." The comparable earnings standard is based on the finding of a rate of return similar to other companies of similar risks and nothing more. The comparable earnings test is consistent with the proposition espoused by Dr. Sherman that investors require a greater return to compensate for increased risk. It is not consistent with a procedure in which the necessarily different returns of companies of differing risk are averaged to reach the recommended return. By so doing, Dr. Sherman is following the principle that companies with differing degrees of risk should have similar rates of return. There is nothing in this methodology that prevents the conclusion that Union Electric is entitled to the same return as a high-risk enterprise like Lucky Stores." Furthermore, the inclusion of flotation and market pressure costs not only is totally inappropriate for a company such as Missouri Utilities which will never incur any of these costs, it assumes, without explanation that the socalled "neutral set" of comparable companies do not already have these costs included in their calculated returns. Also, if as Dr. Sherman posits, the risk of Union Electric earning a return of 10.30% is identical to the risk of Lucky Stores earning a return of 30.4% and U.S. Steel earning 2.5%, increasing Union Electric's return by almost 60% would certainly change its risk-return ratio and most likely move Union Electric into an entirely different set. Because the method by which Dr. Sherman determined his set of 54 companies completely departs from the proper definition of comparability, and, in fact, distorts the entire issue, his analysis cannot be considered valid.

Staff witness, William E. Hutt, and Intervenor witness, Edgar H. Bernstein, also made comparative studies. Mr. Hutt compared Missouri Utilities to the returns of six other primarily electric utilities of similar size and equity ratios. He performed a similar analysis of Union Electric and seven other electric utilities chosen by the same criteria. Mr. Hutt looked at aspects of Missouri Utilities' and Union Electric's financial situations, other than recent returns, including costs of debt and preferred stock, interest coverage, amount of construction, and operating ratios. He stated that the fortunes of the two companies were intertwined and that a

determination of the risks and appropriate returns for Missouri Utilities required a study of both. Mr. Hutt concluded that Missouri Utilities was generally in a more favorable position financially and was less risky than the companies in his comparable group. He stated that a favorable average return on common equity and an extremely low percentage of AFUDC as a percentage of return available to common equity were important factors in reaching this finding. Mr. Hutt also concluded that the favorable position of Missouri Utilities was tempered somewhat by the fact that Union Electric was in a relatively poor financial situation. He cited a low average return on common equity, a thin average common equity ratio, a high pay-out ratio, a decline in earnings per share, and an increase in the number of shares outstanding as reasons for this determination. He recommended a return on equity of 12.70%.

Intervenor witness, Edgar H. Bernstein, proposed a return on common equity of 12.5 to 13.0%. His analysis of the cost of capital for Missouri Utilities was based on the use of earnings-price ratios for Missouri Utilities parent company, Union Electric, and for a selected group of companies allegedly comparable to Missouri Utilities. Data was compared for 1968 through 1976. The comparable companies had the following characteristics: (1) 1976 revenues between 25 and 65 million dollars, (2) at least 75% of their total operating revenues derived from the sales of electricity, (3) stock market prices available for the period selected, (4) complete financial data for the period selected, and (5) state-wide regulation of the operating companies on an original cost basis. Four companies met this criteria.

Witness Bernstein found that for the period 1968 through 1976 the average earnings-price ratio was 10.02% for Union Electric and 10.20% for his list of comparable companies. He then made a 10% adjustment for market pressure and flotation. Dr. Bernstein made an additional upward adjustment of 10.4 to 15.9% to allow for the future growth of earnings per share. He arrived at a range of returns of 12.29 to 12.90% for Union Electric and 12.51 to 13.13% for the comparable group. These ranges were the basis of his 12.5 to 13.0% recommendation.

Finally Dr. Bernstein verified his results by looking at the average returns to Union Electric and the comparable companies for the period of his study and by calculating DCF returns for these companies using Dr. Sherman's method but substituting 1976 dividend yield on market price for dividend yield on book value. He contended that these computations confirmed his proposal."

All three return witnesses relied heavily on their studies of the appropriate return on common equity for Union Electric but did not analyze in any depth the extent to which the appropriate return for the jurisdictional portion of Missouri Utilities