1.4. El Nuevo Periodismo estadounidense
1.4.2. Tom Wolfe
Q. PLEASE DESCRIBE CASH WORKING CAPITAL INCLUDED IN RATE 1
BASE. 2
A. Cash working capital is the amount of investor-supplied capital necessary to 3
finance cost of service expenses between the time the expenditures are required 4
to provide the service to customers and the time cash is received for that 5
service. To determine the allowance of cash working capital, the Commission 6
has traditionally accepted the use of a lead-lag study. 7
Q. DID THE COMPANY PERFORM A LEAD-LAG STUDY THAT WAS USED TO 8
DERIVE THE CASH WORKING CAPITAL AMOUNT IN RATE BASE IN THIS 9
CASE? 10
A. Yes. The Company prepared a lead-lag study based on the twelve months 11
ending May 31, 2011, which was used for both the FTY and HTY cost of service 12
studies. The lead-lag study is presented in two exhibits -- Exhibit No. DAB-9 is a 13
summary of the lead-lag study for all components, and Exhibit No. DAB-10 is the 14
detail supporting the study. Exhibit No. DAB-10 is voluminous and is being 15
provided as a CD-ROM. 16
Q. PLEASE DESCRIBE A LEAD-LAG STUDY. 17
A. A lead-lag study is a method used to measure the amount of working capital 18
required to finance a utility’s day-to-day operations. There are two parts in a 19
lead-lag study. First, the expense lead must be calculated. An extensive and 20
detailed study of the payment practices for each cash expense is made by 21
measuring the period of time from when the Company receives goods or 22
68
services and the date the expense is paid (the “service period”). Statistical 1
sampling can be used to determine the expense lead. Once the expenses to be 2
reviewed (census group or sample) have been determined, each invoice is 3
reviewed to determine the service period. The service period’s mid-point date is 4
calculated. Using the check date as the payment date, the mid-point is 5
subtracted from the payment date, resulting in the number of lead days. 6
Second, the revenue lag must be calculated. The revenue lag is the time 7
between the mid-point of the service period to the date when the Company 8
receives payment from its customer. Depending on the number of customers, 9
statistical sampling can be used to determine the revenue lag. 10
The expense lead is then subtracted from the revenue lag to determine 11
the number of days until the Company is compensated for its expense payout. 12
This net number of days is converted to an annual number by dividing by 365 13
days, which is referred to as the cash working capital factor. The cash working 14
capital factor is multiplied by the corresponding test period expense items and 15
then added to rate base. Cash working capital factors can be positive or 16
negative, depending upon whether the expense lead is shorter or longer than 17
the revenue lag. 18
Q. WHAT STATISTICAL SAMPLING METHODOLOGY DID THE COMPANY 19
USE IN THE LEAD-LAG STUDY PERFORMED IN THIS CASE? 20
A. The Company used the same statistical sampling method to calculate the 21
lead-lag study in this case as was used in the electric rate case in Docket No. 22
06S-234EG, which both the Colorado PUC Staff and the Office of Consumer 23
69
Counsel agreed would be used in future studies. 1
Revenue lag parameters – 2
• Confidence level: 95% 3
• Precision: 5% 4
• Proxy mean and variance: mean and variance from the 2008 electric 5
study as a starting point for the sample size calculation. 6
• For sampled data sets: any accounts drawn with records for fewer 7
than eleven months will be discarded and a new account drawn from 8
the sample. 9
• For census or population data sets: all accounts will be used, 10
regardless of the number of records within each account. 11
• Sample size: consistent with the preceding two parameters, an 12
increase in sample size of no less than 15% is required in order to 13
achieve the confidence and precision requirement as stated above, 14
to compensate for incomplete data, incomplete records, and possible 15
distortion in sample size due to use of mean and variance from the 16
2008 electric lead-lag study as a proxy mean and variance in this 17
study. 18
• Sampling: draw without replacement. 19
Expense lead parameters - 20
• Confidence level: 90% 21
• Precision: 10% 22
• Proxy mean and variance: mean and variance from the 2008 electric 23
study as a starting point for sample size calculation. 24
• Sample size: consistent with the preceding two parameters, an 25
increase in sample size of no less than 10% is required in order to 26
achieve confidence and precision requirement as stated above, to 27
compensate for incomplete data, incomplete records, and possible 28
distortion in sample size due to use of mean and variance from the 29
70
2008 electric lead-lag study as a proxy mean and variance in this 1
study. 2
• Stratified sampling/probability proportional to size (“PPS”) sampling: 3
acceptable. 4
• Sampling: draw without replacement. 5
Q. WHAT PROCESS DOES THE COMPANY FOLLOW WHEN PREPARING A 6
LEAD-LAG STUDY FOR A RATE CASE FILING? 7
A. The process used to prepare a lead-lag study for a rate case filing is presented 8
in Exhibit DAB-7. 9
Q. WHAT CASH EXPENSE ITEMS ARE INCLUDED IN THE EXPENSE LEAD 10
CALCULATION? 11
A. The following cash expense items have historically been included in the expense 12
lead calculation, and were included in the study prepared for this case: 13
• Electric coal for steam production; 14
• Natural gas for other power generation; 15
• Oil for electric generation; 16
• Electric purchased power; 17
• Labor Operating and Maintenance (“O&M”) expense; 18
• Non-Labor O&M expense; 19
• XES charges booked to O&M expense; 20
• Incentive pay; 21
• Paid time off; 22
• Taxes other than income taxes, e.g., property tax and payroll taxes; 23
• State income taxes; 24
• Federal income taxes; 25
• Franchise fees paid; and, 26
• Sales taxes paid. 27
71
Q. DID THE COMPANY INCLUDE INTEREST ON LONG-TERM DEBT IN THE 1
EXPENSE LEAD CALCULATION? 2
A. No. Interest on long-term debt is not included in the lead-lag study. The 3
Commission has determined in previous Public Service rate cases in Docket 4
Nos. I&S 1640, 96S-290G, and 09AL-299E that interest on long-term debt 5
should not be included as a component in the cash working capital allowance. 6
This issue was again disputed in the recent gas rate case, Docket No. 10AL- 7
963G, however, the parties reached a settlement agreement in which this issue 8
was not addressed. 9
Q. BRIEFLY EXPLAIN THE PROCEDURES USED TO DETERMINE THE 10
EXPENSE LEAD. 11
A. The Company used statistical sampling to determine the expense lead for the 12
coal for steam production, natural gas for other power generation, oil for electric 13
generation, purchased power, and non-labor O&M cash working capital expense 14
categories. One hundred percent of the invoices and payments were reviewed 15
and service dates gathered for the O&M Labor, and the various tax cash 16
working capital expense categories. The expense lead is the average number of 17
days from the time of service to the date the Company remits payment for the 18
service to the vendor. The expense lead for each invoice is determined by 19
taking the sum of the following periods: 1) the service period, based on the mid- 20
point of each invoice’s service period; 2) the payment period, based on the 21
number of days it takes for the Company to remit payment to the vendor from 22
the mid-point date of each invoice’s service period; and 3) a half day is added to 23
72
bring the payment date to noon of that day. The expense lead days are 1
weighted by the amount of the invoices. 2
Q. HOW DID THE COMPANY CALCULATE THE CASH WORKING CAPITAL 3
ASSOCIATED WITH THE FUEL, PURCHASED ENERGY AND PURCHASED 4
CAPACITY COSTS? 5
A. The Company multiplied the applicable net lead-lag factors by the per-book test 6
period fuel, purchased energy and purchased capacity expenses, instead of the 7
pro forma amounts. Currently, the electric department has no fuel or purchased 8
energy in base rates, as all electric energy costs are recovered through the 9
ECA. Similarly, all purchased capacity costs are recovered through the PCCA. 10
Therefore, using per-book expense is most representative for calculating a cash 11
working capital amount. The following cash working capital items were 12
calculated in this manner: coal for steam production; natural gas for other power 13
generation, oil for generation, and electric purchased power. 14
Q. PLEASE DESCRIBE HOW THE EXPENSE LEAD WAS CALCULATED FOR 15
THE CASH WORKING CAPITAL ITEM RELATING TO THE XES CHARGES 16
TO PUBLIC SERVICE. 17
A. The Company has calculated the cash working capital expense lead for billings 18
from XES to Public Service using the same methodology that has been used in 19
its last several rate cases. XES provides administrative, accounting and legal 20
services to Public Service and other Xcel Energy subsidiaries. The Company 21
pays XES on approximately the 23rd day of the month following the month in 22
which the services were rendered. The expense lead is calculated by adding 23
73
the service period (the mid-point of each month’s service period) to the payment 1
period (the number of days it takes for the Company to remit payment to XES). 2
Q. PLEASE DESCRIBE THE CASH WORKING CAPITAL ALLOWANCE THAT 3
IS ADDED TO RATE BASE TO REIMBURSE XES FOR FINANCING THE 4
PUBLIC SERVICE CHARGES. 5
A. Consistent with the methodology that has been used in its last several rate 6
cases, the Company has calculated a cash working capital factor that is applied 7
to the XES charges to account for the financing costs incurred by XES before 8
they are paid for the services rendered. The revenue lag is the number of days 9
it takes for Public Service to pay for services rendered. The expense lead is the 10
same as those used by Public Service, since both companies have the same 11
accounts payable payment practices. 12
Q. BRIEFLY EXPLAIN THE PROCEDURES USED TO DETERMINE THE 13
REVENUE LAG. 14
A. The revenue lag was calculated using data from the Company’s customer billing 15
system. The Company used statistical sampling for the customers billed under 16
rate schedules with a large number of customers, and used 100% sampling for 17
the customers under rate schedules with less than 600 accounts. The revenue 18
lag was calculated for each invoice. The revenue lag is the average number of 19
days from the time of service to the date the Company receives payment from 20
the customer. The revenue lag is determined by taking the sum of the following 21
periods: 1) the meter-reading period, based on the mid-point of each month’s 22
service period; 2) the collection lag, based on the number of days it takes for the 23
74
customers to pay their bills from the mid-point date of the service period; and 3) 1
an additional half day is added to account for the posting of the customer 2
receipts to the Company’s bank account. An average lag day value for each 3
rate schedule was calculated and weighted with the percent of total revenue. 4
Q. WHAT ARE THE RESULTING LEAD-LAG FACTORS THE COMPANY HAS 5
CALCULATED FOR USE IN DETERMINING CASH WORKING CAPITAL IN 6
THIS CASE? 7
A. The resulting lead-lag factors are presented on Exhibit No. DAB-9, page 1. 8
These cash working capital factors were then weighted by the applicable test 9
period costs, as presented on Exhibit No. DAB-1, Schedule 12 for the FTY cost 10
of service, and Exhibit No. DAB-3, Schedule 12 for the HTY cost of service. The 11
calculation of Cash Working Capital is presented on Exhibit No. DAB-1, 12
Schedule 3, page 6, lines 253 through 297 and Exhibit No. DAB-3, Schedule 3, 13
page 6, lines 253 through 297. 14
IX. COST OF FUEL AND PURCHASED POWER