Loading...
HomeMy WebLinkAboutWWP68AFN.txt 1 SANDPOINT, IDAHO, TUESDAY, JUNE 8, 1999, 1:15 P. M. 2 3 4 COMMISSIONER SMITH: Welcome back, ladies 5 and gentlemen. We'll resume our hearing. I believe 6 we're at the point where Mr. Woodbury is cross-examining 7 Mr. Dukich. 8 9 THOMAS D. DUKICH, 10 produced as a witness at the instance of Avista 11 Corporation, having been previously duly sworn, resumed 12 the stand and was further examined and testified as 13 follows: 14 15 CROSS-EXAMINATION 16 17 BY MR. WOODBURY: 18 Q Good afternoon. 19 A Good afternoon. 20 Q You're the witness supporting the Company's 21 request for a return on equity adder. Is the Commission 22 under any statutory obligation to provide Avista with any 23 more than a just and reasonable return? 24 A Obligation? 25 Q Yes. 184 CSB REPORTING DUKICH (X) Wilder, Idaho 83676 Avista 1 A I think they have the option of doing that 2 on the basis of, I guess, prior Commission orders as far 3 as I understand. It's, I guess, allowed. Whether 4 they're obligated, I'd say probably not, but I'm not 5 familiar enough with the statutes to say that. 6 Q Have you quantified what 25 basis points 7 equates to in dollars per year? 8 A We have. 9 Q And what does that equate to? 10 A It's about a half million dollars a year. 11 Q It's been 12 years since the Company -- 12 A Excuse me, Scott. 13 Q It's been 12 years since the Company's last 14 general rate case for electric. Was the Company at any 15 time since that rate case by Commission order or 16 settlement agreement precluded from filing a rate case 17 with the Commission? 18 A I don't believe so. I don't think we had 19 any rate freezes in effect or anything. Not to my 20 recollection. 21 Q Has it been a similar period of time since 22 you've had a general rate case in Washington? 23 A Yes. 24 Q And have you had natural gas general rate 25 cases in the interim? 185 CSB REPORTING DUKICH (X) Wilder, Idaho 83676 Avista 1 A In Washington we have. 2 Q In Idaho? 3 A You know, I can't recall in Idaho. I don't 4 think we have. 5 Q Do you have any plans to file a general 6 rate case for natural gas in Idaho? 7 A We will look at it, but we don't have any 8 immediate plans. 9 Q Would I be correct in assuming that 10 customer growth, revenue per customer, distribution plant 11 per customer are operating indices specifically reviewed 12 by the Company on a regular basis? 13 A Could you repeat the list? 14 Q Yeah, customer growth, revenue per 15 customer, distribution plant per customer. 16 A I can't say we specifically review those as 17 indices. I think what we do do is track monthly results 18 of operations and I think that's filed with the 19 Commission and we do semi-annual reports on a Commission 20 basis and do that same rate of return calculation on a 21 jurisdictional basis monthly, so to the extent that those 22 are all rolled up into a rate of return number, yes, but 23 I don't think those are the indices that are tracked for 24 profitability, necessarily. 25 Q And the reports that you file with the 186 CSB REPORTING DUKICH (X) Wilder, Idaho 83676 Avista 1 Commission have you been filing since 1986 or has it just 2 been -- 3 A You know, I'd have to check on that, but we 4 certainly have filed them for the last five to ten 5 years. I don't know if it goes all the way back to '86 6 or not. It probably does, so that would be, what, 12 7 times 12, 144 reports. 8 Q Do the Company's exhibits in this case 9 reflect the incremental annual increase in distribution 10 plant per customer? 11 A Yes, I would think they do. They reflect 12 the -- 13 Q For each year? 14 A Oh, each year? Yeah, they would reflect 15 the cumulative change in distribution plant from the last 16 rate case. 17 Q In showing just the beginning year and the 18 final year, are you inferring that the increase was 19 incrementally standard? 20 A No. In fact, I think if you look at my 21 exhibit, you'll find that you don't get a nice, smooth 22 linear movement from one year to the next, necessarily. 23 It's lumpy. That would be true in the generation 24 resource side, too, because that actually grew more than 25 the distribution plant. 187 CSB REPORTING DUKICH (X) Wilder, Idaho 83676 Avista 1 Q Does the Company's power supply cost 2 increase at a similar rate to the CPI or the COLA? 3 A I would say that Mr. Norwood probably would 4 give a better answer, but my guess would be no, it could 5 be more or less. 6 Q Does the Company perceive that it has a 7 responsibility to its customers to ensure that the cost 8 of service does not get too far out of sync for any 9 particular class? 10 A Yeah, I guess there's an obligation for us 11 to try to fairly reflect what costs are to customers. 12 Q But cost of service is just one factor, 13 though, that is looked at in determining a reasonable 14 rate? 15 A Right, right. 16 Q And -- 17 A It's a major one. 18 Q When you speak of unity, you're speaking of 19 full cost of service? 20 A When we speak of unity, we're saying that 21 each and every class earns some rate of return, say, the 22 proposed rate of return on each and every class would 23 earn that return exactly and to the extent it was under 24 that return, it would be less than unity, so it's a 25 comparative measure. 188 CSB REPORTING DUKICH (X) Wilder, Idaho 83676 Avista 1 Q You state that, within your direct 2 testimony that, the proposed monthly increase for the 3 residential class would be approximately for an average 4 customer about $8.88 per month? 5 A Right. 6 Q And you continue in your testimony 7 regarding rate shock and you state that you don't believe 8 that the increase will constitute rate shock. Would you 9 agree that it's a different question as to whether your 10 requested increase is reasonable or justified and whether 11 the effect of the increase will constitute a hardship on 12 any particular customer? 13 A Could you rephrase that? Is there a 14 difference between rate shock and hardship? 15 Q Whether the rate request that you've asked 16 for is reasonable or justified and whether that rate if 17 authorized would create a hardship on customers. Those 18 are two different questions. 19 A Right. We feel it's a very justified 20 increase, but it could create a hardship. 21 Q And you don't dispute the customers that 22 have filed comments indicating that it would create a 23 hardship in their particular circumstances? 24 A No. I reviewed, I think, just about every 25 one of those and I think some of the people go as far as 189 CSB REPORTING DUKICH (X) Wilder, Idaho 83676 Avista 1 to state their actual monthly income level and then 2 compare their utility bill. I do think, though, that 3 there's not quite an understanding that the fixed income, 4 many of the fixed income, customers who have COLAs did 5 get increases of about 40 percent over the same time 6 period. They sometimes misinterpret that as a one-year 7 inflation rate and also misinterpret the inflation rate 8 as being somehow related to the Company's numbers, not a 9 general change in the economy, so I think there are some 10 educational issues there, but all that aside, whether or 11 not an $8.00 increase is a hardship certainly would 12 depend upon the income level of the customer, where they 13 are now and what they could do to mitigate that and those 14 kinds of things. 15 Q You state that the regulatory compact 16 expects management competency of the Company. Do your 17 shareholders expect management competency or something 18 greater? 19 A I suppose they expect competency and would 20 like something greater. 21 Q And if you in fact don't excel in a more 22 increasingly competitive industry, aren't your jobs at 23 risk? 24 A I suppose, sure. 25 Q So there is some benefit for the Company 190 CSB REPORTING DUKICH (X) Wilder, Idaho 83676 Avista 1 excelling apart from any recompense that you might get 2 from this Commission? 3 A There's a built-in incentive to keep your 4 job, yes. 5 Q Why not just put like a little round-up 6 box, you know, on your billing statement and so if 7 customers think you're deserving of more money, then they 8 can give it to you, you know, sort of voluntarily? You 9 know, they're the ones that are ultimately paying. 10 A I suppose indirectly we do that. I don't 11 know if we actually come out and put a little round-up 12 box there. I think the surveys that have been done show 13 that over 90 percent of the customers rate the Company 14 either as excellent, I think that's about 50 to 55 15 percent, or good in terms of overall satisfaction with 16 the Company if you asked just a generic question. 17 In fact, the ratings have been so high in 18 the excellent category that we changed our benchmarks to 19 reflect only going for the excellent, not the excellent, 20 plus good because we were already well over 90, 95 21 percent, and I think that if you look at the 200 or so 22 customer letters, to my recollection, there are only two 23 that involved complaints about service and one of those 24 was a gas meter that a customer installed and never 25 converted and I can't remember what the other one was, so 191 CSB REPORTING DUKICH (X) Wilder, Idaho 83676 Avista 1 I think overall, despite the rate increase, the 2 customers' rating of the Company is really pretty high. 3 Q And you state that stock prices of 4 well-managed companies are typically bid up by 5 investors. 6 A Yes. 7 Q And it's your contention that you are a 8 well-managed Company? 9 A Yes. 10 Q And yet, Mr. Matthews, your CEO, believes 11 your stock is very undervalued in the market today? 12 A Yes. 13 Q So does that mean your investors don't 14 recognize -- 15 A Not as much as they should. You might talk 16 to Mr. Avera about that as well. 17 Q I think there was some discussion earlier 18 with respect to all of the accolades and awards that 19 you've gotten as a Company, and are there other benefits, 20 tangible or intangible, that accrue to the Company 21 through consistent recognition of accomplishment within 22 the industry? 23 A I'd say yes. 24 Q And what are some of those? 25 A Probably in terms of recruiting. People 192 CSB REPORTING DUKICH (X) Wilder, Idaho 83676 Avista 1 have heard of the Company maybe in disproportion to its 2 size so that recognition for that reason people would 3 want to work for the Company, for instance. It may be 4 that the Company's proposals in different arenas will be 5 given some more credibility than they otherwise might 6 get, portfolio access, for instance. I think the 7 Company's -- I guess we led with our chin up portfolio 8 access which is our approach to having customer choice 9 and did propose that in the region and that ultimately is 10 what the Oregon Commission, for instance, said that the 11 Enron and PG merger should have as part of their 12 resulting company portfolio access approach, so we would 13 take some credit for that being a regional issue, I mean 14 regionally innovated and accepted in Oregon if PG and 15 Enron chooses to accept the merger order. 16 Q You also state that an adjustment to your 17 return on equity is just one of the ways that the 18 Commission can recognize good management and reward 19 shareholders for performance of their management. Can 20 the Commission simply acknowledge your efforts? 21 A Yes. 22 Q And will that have some consequence? 23 A Not as much as the kicker would. We would 24 appreciate any good words, but kickers are good, too. 25 Q Don't you sort of, like, cheapen your 193 CSB REPORTING DUKICH (X) Wilder, Idaho 83676 Avista 1 efforts by expecting payment for your efforts? 2 A A lot of things cross my mind, but I'll 3 simply say no. 4 Q You cite an Idaho Power case as precedent 5 for return on equity adder. Do you know whether Idaho 6 Power in that case specifically asked for an adder or did 7 the Commission just give it to them? 8 A I think they did. I think they asked for a 9 half a percent, as I recall. I have the Order with me, 10 but I won't refer to it if you'll accept my recollection 11 as a half a percent. 12 Q In discussing your demand side management 13 in your rebuttal testimony, you state that it's -- in 14 commenting on Staff's testimony with respect to the 15 accrued balance that the Company maintains in that 16 account, you state that it's necessary and prudent to 17 maintain a positive balance in the conservation account. 18 Are you aware that in the Company's 1994 application in 19 Water Power-E-94-10, attachment D, that the Company's 20 rider implementation point No. 4 specifically states that 21 the DSM tariff rate would also be adjusted up or down to 22 match funding with DSM program costs and to keep the 23 deferred balance as close to zero as possible? 24 A Right. Note that it says as close as is 25 possible and I think that our position here is that the 194 CSB REPORTING DUKICH (X) Wilder, Idaho 83676 Avista 1 experts who are running the program, say, for large 2 commercial accounts, that positive balance is necessary, 3 funding lags, there are large expenditures involved, and 4 also, I think the other thing that it's important to 5 recognize is that the Company, the way the tariff rider 6 works is any money spent in excess is the responsibility 7 of the Company, there's no recovery, so we're cautious, I 8 suppose you might say, because we are at risk for 9 recovery on anything over, but the bottom line is that 10 the DSM people judge that they need the flexibility of 11 having some balance there even though their ultimate goal 12 is to get it as close to zero as possible. They have no 13 intent to carry a balance. The balance is only carried 14 for the sake of running better programs. 15 Q And the average balance that they've 16 maintained is what? 17 A I don't know. I think Mr. Falkner has that 18 number. 19 Q Is it about 750,000? 20 A I think that might be to date. I'm not 21 sure that's the average, though. I don't think that's 22 the average amount and that needs to be spent. We can't 23 keep it or whatever. It rolls over, so it gets spent. 24 Q The Company -- 25 A I might also mention, because I wanted to 195 CSB REPORTING DUKICH (X) Wilder, Idaho 83676 Avista 1 mention this later if it comes up, that in the 2 Commission's Order, and I won't refer to it by page, I 3 think it's page 5 of the November 1998 Order on the DSM 4 extension, that it talked about the Company's proposal 5 for a triple E board which is a board of stakeholders, 6 Commission representation on that board, Commission 7 Staff, as well as other regional experts, I think it's in 8 Staff's testimony, and some of these issues we would hope 9 would be run through the triple E board and the 10 Commission has stated that they consider that a 11 reasonable oversight function and that we would hope that 12 recommendations from the Staff would be run through the 13 triple E board, come to the Commission. 14 Ultimately, clearly, the Commission has the 15 authority and we need to show deference to the Commission 16 decision, but we would just hope that they can be 17 basically run through the triple E board which this 18 Commission has, I guess you might say, blessed as a 19 reasonable oversight group, so issues like this, is that 20 balance appropriate, should it be zeroed out, should it 21 be changed, we would hope it would go to the triple E 22 board first. That would be our proposal. 23 Q Do you recall in the Company's application 24 for extension whether the size of the DSM balance was 25 revealed to the Commission? 196 CSB REPORTING DUKICH (X) Wilder, Idaho 83676 Avista 1 A I don't recall, but I do agree with you 2 that the statement is that the goal is to keep it as 3 close to zero as practically possible and that is a 4 desire. You wouldn't want it to get bigger than you 5 needed it to be to run an effective program, but I don't 6 know if the balance was in the accounting numbers. 7 Q Did you interpret Staff's comments in the 8 extension case as supporting indefinitely a 1.5 percent 9 funding level regardless of the balance maintained? 10 A Indefinitely, no, but since that rate Order 11 is only six months old, it did seem to me to be that it 12 was close enough to the time frame that we've recently 13 had the opportunity to litigate all elements of that 14 program and the Staff was generally supportive, I think 15 only had a few minor comments for cost effectiveness, and 16 we've always agreed that the funding level could be 17 changed. 18 In fact, we have changed it on gas, it's 19 zeroed out on gas, but again I would go back to the 20 triple E board and would hope that if the Staff had 21 concerns about the level of the balance that they'd run 22 it through the triple E board first. 23 Q But the level of the balance wasn't at 24 issue in the extension, it was really a removal of the 25 sunset clause in order to add a perception of tariff 197 CSB REPORTING DUKICH (X) Wilder, Idaho 83676 Avista 1 permanence. 2 A I think if you read through the preamble to 3 the Order, I think that there were probably three or four 4 things stated that were at issue. Extension was only 5 one, funding level was another, triple E board oversight 6 was another, so there was more than just -- the extension 7 was only part of the request, plus it was very explicitly 8 stated in there that consistent funding level over time 9 was highly valuable to any long-term program, I mean you 10 needed to be able to count on a certain funding level, 11 you don't want it to bounce up and down too much, and 12 that's why we've been able to fund DSM through maybe 13 situations where other companies weren't doing that, and 14 I might add that the mechanism itself has turned into 15 kind of a national model. I think Idaho and Washington 16 were the first ones to have it in the United States, I 17 think, and now it's been adopted in California and I 18 think it's being proposed in probably five or six other 19 states. 20 MR. WOODBURY: Thank you, Mr. Dukich, and 21 the Chair, no further questions. 22 COMMISSIONER SMITH: Thank you, 23 Mr. Woodbury. 24 Are there questions from the Commission? 25 COMMISSIONER HANSEN: I just have one. 198 CSB REPORTING DUKICH (X) Wilder, Idaho 83676 Avista 1 COMMISSIONER SMITH: Commissioner Hansen. 2 3 EXAMINATION 4 5 BY COMMISSIONER HANSEN: 6 Q On your rebuttal testimony on page 7, 7 line 7, you say that -- you make the statement that the 8 Company's complaint record is the lowest of Idaho's 9 regulated utility companies. I guess I'm kind of 10 curious, that would probably be compared to the other two 11 major utility companies we regulate in Idaho; is that 12 correct? 13 A Right. I think in Ms. Maxwell's testimony 14 she has an exhibit at the back that shows complaints for 15 the last four years by the electric companies and if you 16 look at the complaints per thousands of customers, I 17 think ours is the lowest. 18 Q So that's based on the total number of 19 complaints, not a percentage according to the number of 20 customers that you have versus the others? 21 A No, I think it's ratioed on a per 1,000 22 customers, so it is on a percentage basis. 23 Q Just one other follow-up. Do you have any 24 comparisons with other regulated companies besides just 25 those in the State of Idaho where you would fall? 199 CSB REPORTING DUKICH (Com) Wilder, Idaho 83676 Avista 1 A You know, we had, we used to -- I'd say 2 about five years ago we did a survey that compared us to, 3 I guess, just in customer perception phone companies, 4 other regulated businesses and we scored high on that and 5 I can't remember the exact numbers, however, so I think 6 our record was also good compared to other regulated 7 businesses nationwide. 8 COMMISSIONER HANSEN: Thank you. 9 COMMISSIONER SMITH: Do you have redirect, 10 Mr. Meyer? 11 MR. MEYER: I do not. Thank you. 12 COMMISSIONER SMITH: Thank you for your 13 help, Mr. Dukich. 14 (The witness left the stand.) 15 MR. MEYER: We'll proceed with Mr. Avera, 16 please. 17 18 19 20 21 22 23 24 25 200 CSB REPORTING DUKICH (Com) Wilder, Idaho 83676 Avista 1 WILLIAM E. AVERA, 2 produced as a witness at the instance of Avista 3 Corporation having been first duly sworn, was examined 4 and testified as follows: 5 6 DIRECT EXAMINATION 7 8 BY MR. MEYER: 9 Q For the record, please state your name and 10 on whose behalf you're appearing today. 11 A William E. Avera and I'm here on behalf of 12 Avista. 13 Q And have you prepared both direct and 14 rebuttal testimony? 15 A Yes, I have. 16 Q Changes to that, sir? 17 A I have one change to my rebuttal. On 18 page 1 at line 13 -- 19 Q Just a moment, let's let everybody catch 20 up. Okay. 21 A I refer to my schedules having been marked 22 for identification as Exhibit 23, I understand they are 23 actually 22. 24 Q Any other changes? 25 A No, sir. 201 CSB REPORTING AVERA (Di) Wilder, Idaho 83676 Avista 1 Q So if I were to ask you the questions that 2 appear in both your direct and rebuttal, would your 3 answers be the same? 4 A They would be. 5 Q Are you also sponsoring what have been 6 marked for identification as Exhibits 5 and 22? 7 A Yes, I am. 8 Q Any changes to those? 9 A No, sir. 10 Q Is the information therein true and correct 11 to the best of your knowledge? 12 A It is. 13 MR. MEYER: With that, I move that the 14 testimony be spread as if read and the exhibits as 15 identified be entered into evidence. 16 COMMISSIONER SMITH: Without objection, it 17 is so ordered. 18 (Avista Corporation Exhibit Nos. 5 & 19 22 were admitted into evidence.) 20 (The following prefiled direct and 21 rebuttal testimony of Mr. William Avera is spread upon 22 the record.) 23 24 25 202 CSB REPORTING AVERA (Di) Wilder, Idaho 83676 Avista 1 I. INTRODUCTION 2 Q. State your name and business address. 3 A. William E. Avera, 3907 4 Red River, Austin, Texas 78751. 5 Q. By whom are you employed and in what 6 capacity? 7 A. I am a principal in Financial Concepts and 8 Applications, Inc. (FINCAP), a firm engaged in financial, 9 economic, and policy consulting to business and 10 government. 11 A. Qualifications 12 Q Describe your educational background, 13 professional qualifications, and prior experience. 14 A I received a B.A. degree with a major in 15 economics from Emory University. After serving in the 16 U.S. Navy, I entered the Ph.D. program in economics at 17 the University of North Carolina at Chapel Hill. Upon 18 graduation, I joined the faculty at the University of 19 North Carolina and taught finance in the Graduate School 20 of Business. I subsequently accepted a position at the 21 University of Texas at Austin where I taught courses in 22 financial management and investment analysis. I then 23 went to work for International Paper Company, Inc. in New 24 York City as Manager of Financial Education, a position 25 in which I had responsibility for all corporate education 203 Avera, Di 1 WWP 1 programs in finance, accounting, and economics. 2 In 1977 I joined the staff of the Public Utility 3 Commission of Texas (PUCT) as Director of the Economic 4 Research Division. During my tenure at the PUCT, I 5 managed a division responsible for financial analysis, 6 cost allocation and rate design, economic and financial 7 research, and data processing systems, and I testified in 8 a number of cases on a variety of financial and economic 9 issues. 10 11 / 12 13 / 14 15 / 16 17 18 19 20 21 22 23 24 25 204 Avera, Di 1A WWP 1 Since forming FINCAP in 1979, I have participated in a 2 wide range of analytical assignments involving 3 utility-related matters on behalf of 4 utilities, industrial customers, municipalities, and 5 regulatory commissions. I have prepared and presented 6 expert witness testimony before a number of regulatory 7 authorities addressing cost of capital, revenue 8 requirements, cost allocation, and rate design issues in 9 the areas of electric, telephone, gas, and water 10 utilities. I have also served as Lecturer in the Finance 11 Department at the University of Texas at Austin, and have 12 taught in the evening graduate program at St. Edward's 13 University for the last eighteen years. I hold the 14 Chartered Financial Analyst (CFA) designation and have 15 served as an officer of various professional 16 organizations and societies. Currently, I am the 17 co-chair of the Synchronous Interconnection Committee, 18 which advises the legislature on the costs and benefits 19 of connecting Texas to the national electric transmission 20 grid. I was appointed to this position by the PUCT, with 21 the Governor's approval. In addition, I serve on the 22 Board of Directors of Georgia System Operations 23 Corporation, the system operator for Oglethorpe Power 24 Corporation, the nations' largest generation and 25 transmission cooperative. A resume containing the 205 Avera, Di 2 WWP 1 details of my experience and qualifications is attached 2 as Appendix A. 3 B. Overview 4 Q. What is the purpose of your testimony? 5 A. The purpose of my testimony is to present 6 to the Idaho Public Utilities Commission (IPUC) my 7 independent assessment of the overall fair rate of return 8 to apply to the original cost rate base of The Washington 9 Water Power Company (WWP). 10 Q. Are you sponsoring any exhibits to be 11 introduced in this 12 13 / 14 15 / 16 17 / 18 19 20 21 22 23 24 25 206 Avera, Di 2A WWP 1 proceeding? 2 A. Yes, I am. My exhibit consists of 6 3 schedules, 2 appendices, and 3 tables. It has been 4 marked for identification as Exhibit No. 5. 5 Q. Please summarize the basis of your 6 knowledge and conclusions concerning the issues to which 7 you are testifying in this hearing. 8 A. I utilized a variety of sources of 9 information in preparing my analyses and testimony in 10 this case that would normally be relied upon by a person 11 in my capacity. In connection with the present filing, I 12 reviewed numerous documents relating to WWP, including 13 bond rating agency reports and financial filings, and had 14 discussions with management regarding the operations, 15 finances, and regulation of WWP. I also examined 16 information relating to capital markets generally and 17 investor perceptions, requirements, and expectations for 18 utilities specifically. These sources, coupled with my 19 experience in the fields of finance and utility 20 regulation, enabled me to acquire a working 21 knowledge of WWP and formed the bases for my conclusions. 22 Q. What is the role of the rate of return in 23 setting a utility's rates? 24 A. The rate of return serves to compensate 25 shareholders for the use of their capital to finance the 207 Avera, Di 3 WWP 1 plant and equipment necessary to provide utility service. 2 Investors only commit money in anticipation of earning a 3 return on their investment commensurate with that 4 from other investment alternatives having comparable 5 risks. Consistent with both sound regulatory economics 6 and the standards specified in the Bluefield (1923) and 7 Hope (1944) cases, the return on investment allowed 8 a utility should be sufficient to: 1) fairly compensate 9 capital invested in the utility, 2) enable the utility to 10 offer a return 11 12 / 13 14 / 15 16 / 17 18 19 20 21 22 23 24 25 208 Avera, Di 3A WWP 1 adequate to attract new capital on reasonable terms, and 2 3) maintain the utility's financial integrity. 3 Q. How did you go about developing a fair rate 4 of return for WWP? 5 A. My evaluation began with a review of the 6 operations and finances of WWP, along with general 7 conditions in the electric utility industry. With this 8 as a background, I next discussed the mix of investor 9 supplied capital -- debt, preferred securities, and 10 common equity -- used as weightings to calculate the 11 overall rate of return. Average costs of debt and 12 preferred securities were then calculated, and 13 various analyses were conducted to determine a fair rate 14 of return on common equity. These included discounted 15 cash flow (DCF) analyses applied to a group of publicly 16 traded electric utilities, as well as risk premium 17 methods encompassing alternative approaches and studies. 18 Finally, the findings of these analyses were combined to 19 calculate an overall rate of return applicable to WWP's 20 original cost rate base. 21 C. Summary of Conclusions 22 Q. What capital structure has WWP requested in 23 this case? 24 A. WWP's requested capital structure is based 25 on outstanding debt, preferred securities, and common 209 Avera, Di 4 WWP 1 equity balances at December 31, 1997, adjusted for known 2 and measurable changes through September 30, 1998. WWP 3 elected to reduce common equity by the amount of its 4 equity investment in certain non-regulated subsidiaries, 5 resulting in a capitalization composed of approximately 6 48.3 percent long-term debt, 4.0 percent short-term debt, 7 10.6 percent preferred securities, and 37.1 percent 8 common equity. 9 Because the equity ratio implied by WWP's 10 requested capital structure falls outside the range 11 maintained by the other electric 12 13 / 14 15 / 16 17 / 18 19 20 21 22 23 24 25 210 Avera, Di 4A WWP 1 utilities used to estimate the cost of equity, it 2 suggests greater financial risk than the proxy group. 3 Indeed, after adjusting for WWP's relatively greater 4 reliance on preferred stock, the total debt ratio 5 associated with its requested capital structure exceeds 6 the upper benchmark required to maintain a single-A bond 7 rating. 8 Q. How were the costs assigned to debt and 9 preferred securities determined? 10 A. The costs associated with the long- and 11 short-term debt components of WWP's capital structure 12 reflect embedded interest rates, adjusted for 13 amortization of capitalized issuance costs over the 14 average term of the respective issues. Similarly, the 15 costs of WWP's preferred securities were based on the 16 dividend yield of each series and also included 17 amortization of related issuance expense. 18 Q. What are your findings regarding the cost 19 of equity? 20 A. The electric utility industry is in the 21 midst of dramatic structural change, and conventional 22 applications of the constant growth DCF model currently 23 do not capture investors' long-term expectations 24 associated with increasing competition, diversification, 25 and consolidation in the industry. Therefore, I relied 211 Avera, Di 5 WWP 1 on a non-constant form of the DCF model to estimate the 2 cost of equity for WWP. 3 My multi-stage DCF model was applied to a proxy 4 group of 8 other single-A rated electric utilities, and 5 produced cost of equity estimates falling in the 11.1 to 6 11.8 percent range, and averaging 11.5 percent. 7 Meanwhile, risk premium analyses based on alternative 8 approaches and studies suggested a cost of equity range 9 for a single-A rated electric utility of 10.75 to 12.25 10 percent. 11 12 / 13 14 / 15 16 / 17 18 19 20 21 22 23 24 25 212 Avera, Di 5A WWP 1 Q. What is your recommended fair rate of 2 return on common equity for WWP? 3 A. Based on these findings, I concluded that 4 common stock investors presently require a rate of return 5 from WWP in the range of 11.0 to 12.0 percent. This 6 range encompasses the 11.1 to 11.8 percent DCF cost of 7 equity range and overlaps all but the lower and upper 8 ends of the 10.75 to 12.25 percent range indicated by my 9 risk premium analyses. 10 This "bare bones" cost of equity range, however, 11 does not recognize flotation costs incurred in connection 12 with past and future sales of common stock, nor does it 13 incorporate any allowance to account for the greater 14 financial risk implied by WWP's requested capital 15 structure. Because the precise calculation of the 16 adjustments for these factors is problematic, I added a 17 minimal adjustment of 25 basis points, producing a fair 18 rate of return on equity range for WWP of 11.25 to 12.25 19 percent. I recommended that WWP be authorized a fair 20 rate of return on equity at the midpoint of this range, 21 or 11.75 percent. Please note that this recommendation 22 does not explicitly incorporate any allowance for WWP's 23 exemplary performance or efficient and economic 24 management discussed in the testimony of Mr. Thomas D. 25 Dukich. 213 Avera, Di 6 WWP 1 Q. What overall rate of return do you 2 recommend be applied to WWP's rate base? 3 A. Combining WWP's requested capital structure 4 with the respective costs of each component, including a 5 12.0 percent cost of equity which incorporates the 6 incentive return recommended by Mr. Dukich, resulted in 7 an overall rate of return on WWP's invested capital of 8 9.446 percent. 9 10 / 11 12 / 13 14 / 15 16 17 18 19 20 21 22 23 24 25 214 Avera, Di 6A WWP 1 II. FUNDAMENTAL ANALYSES 2 Q. What is the purpose of this section? 3 A. As a predicate to subsequent quantitative 4 analyses, this section briefly reviews WWP operations and 5 finances. In addition, it examines the risks and 6 prospects for the electric utility industry as a 7 whole, along with the markets for debt and equity capital 8 and the general economy. An understanding of these 9 fundamental factors driving the risks and prospects of 10 electric utilities is essential in developing an informed 11 opinion of investors' expectations and requirements, 12 which are the bases of a fair rate of return on common 13 equity. 14 A. The Washington Water Power Company 15 Q. Briefly describe WWP. 16 A. Headquartered in Spokane, Washington, WWP 17 is primarily engaged in the electric and natural gas 18 utility business within a 26,000 mile area of northern 19 Idaho and eastern Washington, with gas distribution 20 service also being provided in northeast and southwest 21 Oregon and in the South Lake Tahoe region of California. 22 WWP's operations are organized into three lines of 23 business. The Energy Delivery business includes 24 retail electric and natural gas distribution and 25 transmission services, with the Generation and Resources 215 Avera, Di 7 WWP 1 division being comprised of the generation and 2 production, resource optimization, electric and natural 3 gas commodity trading, and wholesale marketing functions. 4 Meanwhile, Avista Corp., WWP's wholly-owned subsidiary, 5 is engaged in non-regulated activities, including energy 6 commodity trading and marketing as well as investments in 7 real estate and other non-energy businesses. 8 As of December 31, 1997, WWP had total assets of 9 approximately $2.4 billion, with operating revenues 10 totalling $1.3 billion for the most 11 12 / 13 14 / 15 16 / 17 18 19 20 21 22 23 24 25 216 Avera, Di 7A WWP 1 recent fiscal year. The Energy Delivery and Generation 2 and Resources divisions, which comprise WWP's regulated 3 electric and natural gas utility operations, accounted 4 for 94 percent of operating income during 1997. At 5 calendar year-end, WWP's utility operations employed 6 approximately 1,467 individuals. 7 Q. Please describe WWP's electric utility 8 operations. 9 A. WWP provides retail electric service to 10 approximately 297,000 customers, with principal 11 industries in the area including agriculture, mining, and 12 forestry, as well as health care, electronic and other 13 manufacturing, and tourism. Approximately 44 percent of 14 1997 retail electric revenues were from residential 15 customers, with 39 percent from commercial, and 16 16 percent from industrial users. During 1997, WWP sold 17 over 24 billion kilowatt hours (kwh), of which 32 percent 18 were to retail customers, with the remainder being made 19 under wholesale agreements with non-affiliated utilities. 20 With a combined capacity of approximately 1,688 21 Megawatts (MW), WWP's generating facilities include 9 22 hydroelectric generating stations (972 MW) located in 23 Idaho, Montana, and Washington, a 15 percent interest in 24 the Colstrip and Centralia coal-fired plants (423 MW), 25 and two natural gas-fired facilities (245 MW) used 217 Avera, Di 8 WWP 1 primarily to meet peak demand. WWP also owns a 2 wood-waste-fired plant with a generating capacity of 3 approximately 48 MW. In addition, purchased power and 4 exchanges provided approximately 70 percent of WWP's kwh 5 requirements in 1997, reflecting increased sales at 6 wholesale. 7 The electrical output of WWP's hydroelectric 8 plants, which has a significant effect on total energy 9 costs, is dependent on streamflows. Under average 10 operating conditions, WWP expects to meet approximately 11 12 / 13 14 / 15 16 / 17 18 19 20 21 22 23 24 25 218 Avera, Di 8A WWP 1 one-third of its total energy requirements with 2 hydroelectric generation. In contrast to the poor 3 streamflows experienced early in the decade, streamflows 4 were 172 percent, 145 percent, and 120 percent of normal 5 during 1997, 1996, and 1995, respectively. 6 WWP's electric rates and operations are subject to 7 regulation by the respective state agencies in Idaho and 8 Washington, and at the federal level by the Federal 9 Energy Regulatory Commission (FERC). Additionally, all 10 but one of WWP's hydroelectric facilities are subject to 11 licensing under the Federal Power Act, which is 12 administered by FERC. Two of the eight licenses expire 13 in 2001, while five come up for renewal in 2007. 14 Relicensing is not automatic under federal law, and WWP 15 must demonstrate that it has operated its facilities in 16 the public interest, which includes adequately addressing 17 environmental concerns. 18 Q. How are fluctuations in WWP's operating 19 expenses caused by varying hydro conditions accommodated 20 in its rates? 21 A. The IPUC, which regulates approximately 30 22 percent of WWP's retail revenues, has approved a power 23 cost adjustment (PCA) mechanism under which rates are 24 adjusted to reflect changes in power production and 25 supply costs arising from variability in hydroelectric 219 Avera, Di 9 WWP 1 generation. No similar mechanism has been implemented in 2 the Washington jurisdiction. 3 Q. What bond ratings are assigned to WWP's 4 long-term debt? 5 A. WWP's senior debt is presently rated "A3" 6 by Moody's Investors Service (Moody's), with Standard & 7 Poor's Corporation (S&P) and Duff & Phelps Credit Rating 8 Co. assigning these securities an "A" rating. 9 B. Electric Utility Industry 10 Q. What are the general conditions in the 11 electric power industry? 12 13 / 14 15 / 16 17 / 18 19 20 21 22 23 24 25 220 Avera, Di 9A WWP 1 A. Following rapidly escalating electricity 2 prices during the late 1970s and early 1980s, electric 3 utilities and their customers benefited from an extended 4 period characterized by lower fuel costs and relatively 5 low inflation and interest rates. Recently, however, 6 these general economic factors have been overshadowed by 7 structural changes in the electric utility industry 8 resulting from market forces, decontrol initiatives, and 9 judicial decisions. 10 Q. Please describe these structural changes. 11 A. Competition is being increasingly promoted 12 at the federal and state levels. The National Energy 13 Policy Act of 1992, which reformed the Public Utility 14 Holding Company Act of 1935, greatly increased 15 prospective competition for the production and sale of 16 power at the wholesale level. In April 1996, the FERC 17 adopted Order No. 888, which mandated open access to the 18 wholesale transmission facilities of jurisdictional 19 electric utilities. Wholesale wheeling provides 20 transmission-dependent electric utilities with additional 21 energy supply options, but it also has introduced new 22 risks to participants in the wholesale power markets. 23 While policies affecting competition in the 24 electric power industry vary widely at the state level, 25 more than 45 states are considering legislation to 221 Avera, Di 10 WWP 1 implement market reforms and address issues surrounding 2 deregulated pricing, recovery of stranded investment, and 3 restrictions to market entry. Similarly, while a 4 comprehensive restructuring plan has not been implemented 5 in Idaho, the legislature has directed the IPUC to 6 collect cost information on generation, transmission, and 7 distribution, and an interim committee was established to 8 examine restructuring issues. A restructuring committee 9 formed by the governors 10 11 / 12 13 / 14 15 / 16 17 18 19 20 21 22 23 24 25 222 Avera, Di 10A WWP 1 of Idaho, Montana, and Washington is calling for customer 2 choice to be offered by July 1, 1999 and for utilities to 3 functionally disaggregate their operations. Federal 4 legislators are also considering new laws designed to 5 clarify the boundaries between state and federal 6 jurisdiction while hastening the onset of retail 7 competition. At the same time, customer choice 8 initiatives are now being implemented in a number of 9 jurisdictions, including California, Massachusetts, 10 Michigan, New Hampshire, Pennsylvania, and Rhode Island. 11 Q. What are the effects of these structural 12 changes on electric utilities? 13 A. As a result of deregulation and ensuing 14 competition on both the demand and supply sides of the 15 industry, electric utilities' traditional monopoly status 16 is eroding. S&P recognized the mounting business risk of 17 the electric utility industry by adopting in November 18 1993 more stringent standards for the financial ratio 19 guidelines it uses in rating the debt of electric 20 utilities. More recently, in its Industry Surveys: 21 Electric Utilities (February 19, 1998) S&P noted that: 22 The electric power industry is in the midst of a radical change. The monopolistic, tightly 23 regulated utilities created under trust-busting legislation more than 60 years ago are slowly 24 being exposed to competition, particularly in the generation and wholesale power markets. 25 Technological advances and the increased desire 223 Avera, Di 11 WWP 1 for customer choice are spurring the demand for new legislation. Whereas in the past electricity 2 markets were strictly delineated by geographic lines, utilities now have the freedom to cross 3 into one another's territories. (p. 8) 4 Similarly, in its Electric Utilities Industry research 5 report (June 24, 1996) Merrill Lynch recognized that: 6 The electric utility industry is in a monumental transition state at the current time. The 7 transition is from a vertically-integrated, monopoly industry to one that we expect to be very 8 competitive and significantly restructured. We 9 10 / 11 12 / 13 14 / 15 16 17 18 19 20 21 22 23 24 25 224 Avera, Di 11A WWP 1 expect all utility customers to have competitive choices in the next 5-10 years. We expect 2 companies to realign and/or disaggregate their businesses -- some may exit the generation 3 business, others may exit the distribution business -- as well as merge to create larger 4 companies. The risk profile of the electric utility industry is clearly reaching higher levels 5 than it has experienced in the past and will further increase... (p. 3) 6 And in an August 26, 1998 research report A.G. Edwards 7 observed that: 8 As the electric utility industry restructures and 9 adapts to competitive power generation markets, investors must be cognizant of the impact on 10 utility securities. Traditionally a relatively low-risk source of income, many utility companies 11 are adopting more growth-oriented strategies which require greater capital investment. In order to 12 finance growth, many utilities are seeking to reduce the dividend payout ratio or utilize 13 greater leverage. These actions, in addition to the deregulation of the industry, could increase 14 investment risk. (p. 5) 15 Indeed, in August 1998, WWP announced a dividend 16 restructuring plan which resulted in a reduction of the 17 common dividend from $1.24 per share to $0.48 per share, 18 effective with the December 1998 payment. WWP's decision 19 to cut common dividends reflects an industry trend to 20 align financial policies more closely with the 21 competitive sector of the economy, such as industrial 22 firms, and is consistent with an environment 23 characterized by expectations of heightened risk and 24 higher growth opportunities. 25 Q. Are all of the risks associated with the 225 Avera, Di 12 WWP 1 restructuring of the electric industry known at this 2 time? 3 A. No. My experience with deregulation in the 4 transportation and natural gas industries demonstrates 5 that the structural changes associated with deregulation 6 produce consequences that no one can predict. In 7 particular, as prices become primarily market-driven, 8 future changes in prices become inherently uncertain. 9 Much of this uncertainty simply reflects the 10 superior ability of 11 12 / 13 14 / 15 16 / 17 18 19 20 21 22 23 24 25 226 Avera, Di 12A WWP 1 markets to adjust continually both to changing customer 2 needs and to the changing costs of meeting those needs. 3 This point was succinctly stated in the 1997 Economic 4 Report of the President: 5 An insufficiently appreciated property of markets is their ability to collect and distribute 6 information on costs and benefits in a way that enables buyers and sellers to make effective, 7 responsive decisions. ... As tastes, technology, and resource availability change, market prices 8 will change in corresponding ways, to direct resources to the newly valued ends and away from 9 obsolete means. It is simply impossible for governments to duplicate and utilize the massive 10 amount of information exchanged and acted upon daily by the millions of participants in the 11 marketplace. (p. 191) 12 While competition in the electric industry will provide 13 many benefits for both producers and consumers of 14 electric power, participants will become exposed to many 15 uncertainties that regulators have heretofore shielded 16 them from, such as the threat of new entrants and price 17 volatility in the wholesale power markets. On the 18 benefits side, this country's experience with other 19 industries shows that deregulation should enable the 20 electric utility industry to achieve new efficiencies. 21 The 1997 Economic Report of the President cited a recent 22 study that assessed the long-run benefits of 23 deregulation. According to the study, because of 24 innovations stimulated by competition, costs fell 24 25 percent in the airline industry, 30 to 35 percent in the 227 Avera, Di 13 WWP 1 trucking industry, 50 percent in the railroad industry, 2 and 35 percent in the natural gas industry. 3 These changes were impossible to predict at the 4 time each of the industries was deregulated. As more 5 than one wag has put it, the effect of the Staggers Act 6 on railroading was, well, staggering. Revenue per 7 ton-mile, the common gauge of rail prices, declined 16 8 percent between 1980 and 1995. Most notably, with the 9 technological improvements 10 11 / 12 13 / 14 15 / 16 17 18 19 20 21 22 23 24 25 228 Avera, Di 13A WWP 1 stimulated by competition, the railroad industry reduced 2 the total number of locomotives and railcars by 30 3 percent, even as traffic rose 48 percent. To sum it up, 4 the changes wrought by deregulation exceeded the 5 expectations of all knowledgeable industry authorities. 6 Q. Are you aware that Idaho has the lowest 7 electric rates in the country, and that both the IPUC and 8 the state Legislature have publicly expressed doubts 9 about whether deregulation would benefit Idaho consumers? 10 A. Yes. More important than my personal 11 views, however, investors are also following the 12 transition of the nation's electric power markets and 13 recognize that some states, particularly where electric 14 rates are relatively high, have moved faster than others 15 to introduce competition. They are also aware that the 16 pace of legislative action is difficult to predict, and 17 that the industry transition is being driven by 18 broad-reaching considerations such as evolving technology 19 and increasing consolidation, not just by the actions of 20 state legislatures and regulators. Consistent with 21 investors' concern for the "big picture", investment 22 advisory services have noted the actions WWP has taken to 23 transform its operations and financial policies in 24 response to the changing environment faced by utilities 25 throughout the U.S., including Idaho. 229 Avera, Di 14 WWP 1 Q. Are the risks associated with structural 2 change the only ones being faced by the electric utility 3 industry? 4 A. No. Apart from these structural changes, a 5 number of electric utilities, once considered the paragon 6 of financial stability, are in difficult financial 7 straits, with some having even faced bankruptcy. In part 8 to avoid the risks associated with building additional 9 base-load 10 11 / 12 13 / 14 15 / 16 17 18 19 20 21 22 23 24 25 230 Avera, Di 14A WWP 1 generating capacity, electric utilities have pursued a 2 variety of options, such as increased reliance on power 3 purchases from wholesale suppliers and non-utility 4 generators, although these entail additional risks in and 5 of themselves. Meanwhile, many electric utilities have 6 begun to diversify into unrelated activities and expand 7 through mergers and acquisitions. 8 The industry continues to face the risks and 9 uncertainties inherent in operating electric utility 10 systems. Electric utilities are facing increased 11 environmental pressures, such as the Environmental 12 Protection Agency's efforts to force reductions of 13 nitrous oxides to address ozone problems in the 14 Northeastern U.S., lower carbon dioxide emissions which 15 might be mandated in response to the issue of "global 16 warming", as well as measures which may be adopted to 17 protect endangered or threatened fish and wildlife 18 species. These programs could impose significant 19 restrictions on utilities that rely on coal- and 20 hydro-powered generation. Nuclear risk persists for 21 those utilities involved in nuclear plants, although the 22 exposure has largely shifted from construction to 23 operating and decommissioning uncertainties, and 24 utilities remain exposed to the impacts of extreme 25 weather, such as the ice storm damage experienced on 231 Avera, Di 15 WWP 1 WWP's distribution system in 1996. 2 C. Economy and Capital Markets 3 Q. What has been the pattern of interest rates 4 during the 1980s and 1990s? 5 A. After peaking at 16.89 percent in September 6 1981, the average yield on long-term public utility bonds 7 generally fell through 1986, reaching 8.77 percent in 8 January 1987. After climbing during 1988, yields 9 gradually declined to 7 percent in October 1993, and then 10 11 / 12 13 / 14 15 / 16 17 18 19 20 21 22 23 24 25 232 Avera, Di 15A WWP 1 subsequently rose to 9 percent in November 1994. 2 Interest rates fell through January 1996, and continued 3 to decline beginning in May 1997, with investors 4 presently requiring approximately 6.9 percent from 5 average long-term public utility bonds. Long-term 6 average public utility bond rates, the monthly average 7 prime rate, and inflation as measured by the Consumer 8 Price Index ("CPI") since 1979 are plotted in the 9 following graph: 10 11 12 13 14 15 (Graph contained in hard copy of transcript.) 16 17 18 19 20 21 22 23 24 Q. How has the market for common equity 25 capital performed over this same period? 233 Avera, Di 16 WWP 1 A. The last 18 years witnessed the longest 2 bull market in U.S. history, which is generally 3 attributed to low inflation and interest rates, sustained 4 economic growth, a favorable business climate, and 5 widespread merger and acquisition activity. Since 1979, 6 common stocks have, on average, increased over ten times 7 in value, even after accounting for the October 1987 and 8 1989 stock market crashes and the 9 10 / 11 12 / 13 14 / 15 16 17 18 19 20 21 22 23 24 25 234 Avera, Di 16A WWP 1 "corrections" of October 1997 and 1998. While share 2 prices have recently recovered after falling from the 3 all-time highs reached in August 1998, the market remains 4 volatile. The following graph plots the performances of 5 the Dow-Jones Industrial Average, S&P's 500 Composite 6 Index, and New York Stock Exchange Utility Index since 7 1979 (the latter two indices were scaled for 8 comparability): 9 10 11 12 13 14 (Graph contained in hard copy of transcript.) 15 16 17 18 19 20 21 Q. What is the outlook for the U.S. economy 22 and capital markets? 23 A. There are persistent concerns over how long 24 the economic expansion, which began in the latter half of 25 1991, can be sustained, and that a downturn in the U.S. 235 Avera, Di 17 WWP 1 economy is inevitable. While numerous economic 2 indicators suggest that the U.S. economy is continuing to 3 grow, there are increasing signs that the pace of 4 expansion may moderate going forward. These factors 5 cause the economic outlook to remain tenuous, with 6 persistent stock and bond price volatility providing 7 tangible evidence of the uncertainties faced by the U.S. 8 economy. The compound 9 10 / 11 12 / 13 14 / 15 16 17 18 19 20 21 22 23 24 25 236 Avera, Di 17A WWP 1 effect of these economic uncertainties and an 2 increasingly competitive marketplace has caused the risks 3 presently faced by electric utilities probably to be 4 greater now than at any other time in modern history. 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 237 Avera, Di 18 WWP 1 III. CAPITAL STRUCTURE AND 2 COST OF DEBT AND PREFERRED SECURITIES 3 Q. What is the purpose of this section? 4 A. This section discusses the implications of 5 the capital structure on risk and rate of return, and 6 compares the capital structure developed for WWP with 7 those maintained by other electric utilities and against 8 industry benchmarks. In addition, the embedded costs of 9 debt and preferred stock applicable to the respective 10 components of WWP's capital structure are evaluated. 11 A. Principles 12 Q. What is the role of capital structure in 13 setting a utility's rate of return? 14 A. A utility's capital structure reflects the 15 mix of capital -- debt, preferred stock, and common 16 equity -- used to finance its assets. The proportions of 17 a utility's total capitalization attributable to each 18 source of capital are typically used to weight the costs 19 of debt and preferred stock, and rate of return on common 20 equity, in calculating an overall rate of return. 21 Q. Why does this weighting matter? 22 A. The capital structure ratios determine how 23 much weight is given to a particular source of capital 24 and, since the costs of debt and preferred stock and the 25 rate of return on common equity are not the same, this 238 Avera, Di 19 WWP 1 affects the weighted average cost, or overall rate of 2 return, of all sources of capital. 3 Q. Why are the costs of debt and preferred 4 stock, and the rate of return on common equity, not the 5 same? 6 A. The reason for this difference is that 7 debt, preferred stock, 8 9 / 10 11 / 12 13 / 14 15 16 17 18 19 20 21 22 23 24 25 239 Avera, Di 19A WWP 1 and common equity have different characteristics which 2 cause investors to demand a higher rate of return to 3 invest in the common stock of a utility versus loan it 4 money in the form of debt or preferred stock. 5 When investors loan money in the form of debt 6 (e.g., bonds), they enter into a contract whereby the 7 utility agrees to pay the bondholders a specified amount 8 of interest and to repay the principal of the loan in 9 full. The bondholders have a senior claim on available 10 cash flow for these payments, and if the utility fails to 11 make them, they may force it into bankruptcy and 12 liquidation for settlement of unpaid claims. Similarly, 13 when the utility sells investors preferred stock, the 14 utility promises to pay preferred stockholders specified 15 dividends and, typically, to retire the preferred stock 16 on a predetermined schedule. While the rights of 17 preferred shareholders to available cash flow for these 18 payments are junior to creditors, and preferred 19 stockholders cannot compel bankruptcy, their claims are 20 senior to those of common shareholders. 21 The last in line are common shareholders, the 22 residual owners of the utility. They only receive cash 23 flows, if any, that remain after all other claimants -- 24 employees, suppliers, governments, lenders, and preferred 25 stockholders -- have been paid. Therefore, the greater 240 Avera, Di 20 WWP 1 number of investors (i.e., bondholders and preferred 2 stockholders) who have a prior claim on the utility's 3 earnings, the greater the risk to common shareholders. 4 For investors to be willing to bear this additional risk, 5 they require a higher rate of return than lenders and 6 preferred stockholders who have more certain, senior 7 claims on the cash flows of the utility. 8 9 / 10 11 / 12 13 / 14 15 16 17 18 19 20 21 22 23 24 25 241 Avera, Di 20A WWP 1 Q Why doesn't a utility finance itself 2 entirely with debt or preferred stock, since these are 3 less expensive sources of capital? 4 A. If a utility were to attempt to finance 5 itself with 100 percent debt or preferred stock, then 6 there would be no common shareholders. The lenders or 7 preferred stockholders would effectively become the 8 residual owners of the utility, and since they would be 9 exposed to the same risks as if they were common 10 shareholders, they would require a correspondingly higher 11 rate of return as compensation. Accordingly, utilities 12 are generally financed with a mix of debt, preferred 13 stock, and common equity in an effort to produce the 14 lowest overall cost of capital while, at the same time, 15 permitting the utility to maintain its financial 16 integrity and its ability to attract additional capital 17 on reasonable terms. 18 Q. How does the use of greater amounts of debt 19 affect the rates of return required by investors? 20 A. A higher debt ratio, or lower common equity 21 ratio, translates into increased financial risk for all 22 investors. A greater amount of debt, and preferred 23 stock, means more investors have a senior claim on 24 available cash flow, thereby reducing the certainty that 25 each will receive his contractual payments. This, in 242 Avera, Di 21 WWP 1 turn, increases the risks to which lenders and preferred 2 stockholders are exposed, and they require 3 correspondingly higher rates of interest and dividends, 4 respectively, for their risk bearing. From common 5 shareholders' standpoint, higher debt and preferred stock 6 ratios mean that there are proportionately more investors 7 ahead of them, thereby increasing the uncertainty as to 8 the amount of cash flow, if any, that will remain. 9 Again, in accordance with the fundamental risk-return 10 tradeoff principle to be discussed in 11 12 / 13 14 / 15 16 / 17 18 19 20 21 22 23 24 25 243 Avera, Di 21A WWP 1 greater detail later, common shareholders require a 2 correspondingly higher rate of return to compensate them 3 for bearing the greater financial risk associated with a 4 lower common equity ratio. 5 Q. What implications does the transition to 6 competition have for the capital structures maintained by 7 electric utilities? 8 A. The heightened business risks imposed by 9 the evolution to competitive markets will force electric 10 utilities to adopt a more conservative financial posture 11 if credit ratings are to be maintained, as Moody's noted 12 in its Credit Risk Commentary (July 29, 1996: 13 "The key issue," say the analysts in a recent special comment, "is that the competitive 14 industries have much lower operating and financial leverage, and that utilities must streamline both 15 in order to be effective competitors." Analysts say the utilities must do this in order to post 16 stronger financial indicators and maintain their current ratings level ... (p. 3) 17 18 Accordingly, the challenges imposed by evolving 19 structural changes in the industry imply that electric 20 utilities, including WWP, will be required to incorporate 21 relatively greater amounts of equity in their capital 22 structures. 23 B. Capital Structure Ratios 24 Q. What are the sources of capital used to 25 finance WWP's investment in utility plant? 244 Avera, Di 22 WWP 1 A. As shown on Schedule WEA-1, after updating 2 for known and measurable changes through September 30, 3 1998, WWP's capital structure consists of approximately 4 $657.8 million in long-term debt, $54.2 million in 5 short-term debt, and $145.0 million in preferred 6 securities, including $110 million in Preferred Trust 7 Securities. WWP's common equity amounted to 8 approximately $753.4 million at September 30, 1998. 9 Consistent with past cases before the IPUC, WWP elected 10 to reduce total 11 12 / 13 14 / 15 16 / 17 18 19 20 21 22 23 24 25 245 Avera, Di 22A WWP 1 common equity on a consolidated basis to reflect its 2 investment in non-regulated subsidiaries, resulting in an 3 adjusted common equity balance of $512.5 million. As 4 shown on Schedule WEA-1, this produced a capital 5 structure for WWP consisting of 48.030 percent long-term 6 debt, 3.958 percent short-term debt, 10.587 percent 7 preferred securities (including 8.080 percent Preferred 8 Trust Securities), and 37.424 percent common equity. 9 Q. Can an analysis of the capital structures 10 maintained by comparable firms provide a proper basis for 11 evaluating a capital structure? 12 A. Yes. It is generally accepted that the 13 norms established by comparable firms provide a valid 14 benchmark against which to evaluate the reasonableness of 15 a utility's capital structure. The capital structures 16 maintained by similar companies should reflect their 17 collective efforts to finance themselves so as to 18 minimize capital costs while preserving their financial 19 integrity and ability to attract capital. Moreover, 20 these industry capital structures should also incorporate 21 the requirements of investors, both debt and equity, as 22 well as the influence of regulators. 23 Q. What capitalization ratios are maintained 24 by other electric utilities? 25 A. Schedule WEA-2 displays capital structure 246 Avera, Di 23 WWP 1 data for the same group of 8 single-A rated electric 2 utilities used subsequently to estimate the cost of 3 equity. The average capital structure ratios at year-end 4 1997 for this group of electric utilities are shown 5 below: 6 7 / 8 9 / 10 11 / 12 13 14 15 16 17 18 19 20 21 22 23 24 25 247 Avera, Di 23A WWP 1 Capital Component % of Total 2 Long-term Debt 45.2% Preferred Stock 5.5% 3 Common Equity 49.3% 4 Total 100.0% 5 A review of Schedule WEA-2 also reveals that the 6 individual equity ratios for the 8 companies in the group 7 range from a low of 44.5 percent to a high of 55.7 8 percent. 9 Q. How do these ratios compare with other 10 widely cited benchmarks for electric utilities? 11 A. S&P routinely publishes financial ratio 12 guidelines consistent with specific bond ratings. Viewed 13 in conjunction with a utility's business risk profile, 14 S&P's guideline financial ratios for a given rating 15 category (e.g., single-A) vary with the business or 16 operating risk of the company. In other words, a utility 17 with a "Below Average" business position (i.e., higher 18 business risk) would presumably employ less financial 19 leverage to maintain the same credit rating. Currently, 20 S&P has assigned WWP a business profile ranking of "3", 21 which corresponds to an "Above Average" business 22 position. 23 In order to maintain a single-A rating, S&P 24 requires a ratio of total debt-to-total capital in the 41 25 to 52 percent range, depending on a utility's business 248 Avera, Di 24 WWP 1 position, with a median of 47 percent. For firms with an 2 "Above Average" business position, such as WWP, total 3 debt should comprise no greater than 52 percent of total 4 capital. 5 Q. How does WWP's requested capital structure 6 compare with these industry benchmarks? 7 A. Even setting aside the impact of short-term 8 debt on WWP's capitalization ratios, WWP's 48 percent 9 long-term debt ratio is higher 10 11 / 12 13 / 14 15 / 16 17 18 19 20 21 22 23 24 25 249 Avera, Di 24A WWP 1 than the approximately 45 percent average maintained by 2 other single-A rated electric utilities. Similarly, the 3 10.6 percent of capital made up of preferred securities 4 is considerably greater than the roughly 5.5 percent 5 average for the electric utility group. Finally, WWP's 6 37 percent equity ratio falls outside the range of values 7 maintained by the group of other electric companies and 8 is well below the benchmark average of approximately 49 9 percent. 10 Meanwhile, WWP's requested capital capitalization 11 cannot be evaluated directly against S&P's benchmark 12 ratios. As S&P observed in a September 1997 Utility 13 Credit Report: 14 ... leverage as a percent of capital is diminished by WWP's high use of preferred stock in its 15 capital structure. Furthermore, auction rate preferred and preferred trust securities (a 16 quasi-equity instrument) are considered by Standard & Poor's to have less equity attributes 17 than perpetual preferred stock. ... Standard & Poor's methodology is that up to 10% of preferred 18 stock can receive equity treatment (the 10% ceiling representing only perpetual preferred 19 stock). Lower levels are regarded as equity when less permanent preferred vehicles are used. (p. 9) 20 Accordingly, S&P concluded that: 21 Given the nature of WWP's preferred instruments, 22 Standard & Poor's would view the range of 6% to 7% appropriate as being afforded "equity treatment". 23 (p. 9) 24 As shown below, adjusting WWP's capitalization (Schedule 25 WEA-1) to reflect equity treatment of preferred 250 Avera, Di 25 WWP 1 securities equal to 7 percent of total capital results in 2 a total debt ratio of approximately 55.6 percent: 3 4 / 5 6 / 7 8 / 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 251 Avera, Di 25A WWP 1 Capital Component % of Total 2 Long-term Debt 48.0% Short-term Debt 4.0 3 S&P Adjustment to Preferred 3.6 Total Debt 55.6% 4 Preferred Securities 10.6% 5 S&P Adjustment to Preferred (3.6) Adjusted Preferred 7.0% 6 Common Equity 37.4% 7 Total 100.0% 8 9 Thus, the total debt ratio implied by WWP's requested 10 capital structure falls outside S&P's 41 to 52 percent 11 benchmark range for a single-A bond rating, even for a 12 company with an "Above Average" business position. 13 Q. What capital structure ratios are normally 14 used for ratemaking purposes to weight the cost of each 15 source of capital to arrive at an overall rate of return 16 for a utility? 17 A. It depends. If a utility's capital 18 structure is generally in line with industry standards, 19 it is customary to use the utility's actual capital 20 structure ratios to calculate its overall rate of return. 21 On the other hand, when a utility's capital structure 22 falls outside of industry norms, it is not uncommon for 23 regulators to base the utility's rate of return on 24 industry capital structure ratios. By using an industry 25 capital structure, rates reflect the reasonable capital 252 Avera, Di 26 WWP 1 costs of providing a particular service. As noted 2 earlier, the capital structures maintained by 3 stand-alone, publicly traded companies also reflect the 4 capital markets' perceptions of the business risks faced 5 by the industry, and the mix of debt and equity required 6 to accommodate these risks. Finally, using capital 7 structure ratios consistent with industry standards 8 permits the rate of return to be based directly on 9 investors' required rate of return from other electric 10 utilities and 11 12 / 13 14 / 15 16 / 17 18 19 20 21 22 23 24 25 253 Avera, Di 26A WWP 1 avoids any adjustment to account for financial risk 2 differences. 3 Q. What are the implications of using WWP's 4 requested capital structure to calculated the overall 5 rate of return? 6 A. WWP's higher debt and preferred stock 7 ratios, and its lower common equity ratio, imply greater 8 financial risk because a larger proportion of available 9 cash flow is subject to senior claims. However, because 10 there is a greater proportion of lower-cost debt and 11 preferred stock, and correspondingly less higher-cost 12 common equity (including associated income taxes), in 13 WWP's requested capital structure, ratepayers benefit 14 from an effective cost of capital that is lower than 15 would be implied by industry norms. 16 C. Cost of Debt 17 Q. What average cost is associated with WWP's 18 long-term debt? 19 A. As indicated earlier, at December 31, 1997, 20 updated for known and measurable changes through 21 September 30, 1998, WWP had approximately $657.8 million 22 in long-term debt outstanding. This balance is composed 23 of first mortgage bonds, secured medium term notes, 24 Kettle Falls pollution control bonds, and medium term 25 notes, with the interest rates attributable to each 254 Avera, Di 27 WWP 1 specific issue being detailed in Schedule WEA-3. Besides 2 interest expense, WWP necessarily incurs various 3 issuance-related costs in connection with securing debt 4 capital. Although these costs are capitalized and 5 amortized over the life of the corresponding debt issue, 6 none is included in WWP's rate base or operating 7 expenses. Accordingly, as shown on page 2 of Schedule 8 WEA-3, combining the annual interest cost for each series 9 of long-term debt outstanding with related issuance costs 10 produced an average cost of long-term debt for WWP of 11 8.011 percent. 12 13 / 14 15 / 16 17 / 18 19 20 21 22 23 24 25 255 Avera, Di 27A WWP 1 Q. What cost rate was assigned to the 2 short-term debt component of WWP's capital structure? 3 A. Also shown on page 2 of Schedule WEA-3, the 4 weighted cost of WWP's short-term debt outstanding 5 equaled 6.255 percent. 6 D. Cost of Preferred Securities 7 Q. What preferred securities does WWP have 8 outstanding? 9 A. As shown on page 7 of Schedule WEA-3, WWP 10 has three series of preferred securities outstanding -- 11 $60 million in Trust Originated Preferred Securities and 12 $50 million in Floating Rate Capital Securities 13 (together, the Preferred Trust Securities) and a $35 14 million issued of cumulative preferred stock. As with 15 its debt, WWP incurred issuance costs in connection with 16 the sale of its preferred securities. As detailed in 17 Schedule WEA-3, including the amortization of these 18 expenses along with the annual dividend cost resulted in 19 a weighted cost of Preferred Trust Securities for WWP of 20 8.113 percent and a cost rate of 8.151 percent for the 21 cumulative preferred stock. 22 23 24 25 256 Avera, Di 28 WWP 1 IV. RATE OF RETURN ON EQUITY 2 Q. What is the purpose of this section? 3 A. This section describes my evaluation of the 4 cost of common equity for WWP. Initially, the concept of 5 the cost of equity is examined, as is the risk-return 6 tradeoff principle fundamental to capital markets. Next, 7 various quantitative analyses are developed as a guide to 8 investors' current required rate of return on equity, 9 including discounted cash flow analyses and alternative 10 risk premium methods applied to a group of electric 11 utilities having investment risks similar to those of 12 WWP. 13 A. Economic Standards 14 Q. What role does the rate of return on common 15 equity play in a utility's rates? 16 A. As indicated earlier, the return on common 17 equity serves to compensate shareholders for the use of 18 their capital to finance the plant and equipment 19 necessary to provide utility service. Investors only 20 commit money in anticipation of earning a return on their 21 investment commensurate with that from other investment 22 alternatives having comparable risks. Moreover, the 23 return on common equity is integral in achieving the 24 sound regulatory, economic, and legal objectives of rates 25 that are sufficient to: 1) fairly compensate capital 257 Avera, Di 29 WWP 1 investment in the utility, 2) enable the utility to offer 2 a return adequate to attract new capital on reasonable 3 terms, and 3) maintain the utility's financial integrity. 4 Q. How is a fair rate of return on common 5 equity customarily determined? 6 A. Conventional regulation is largely based on 7 the cost of 8 9 / 10 11 / 12 13 / 14 15 16 17 18 19 20 21 22 23 24 25 258 Avera, Di 29A WWP 1 providing service. Common equity, like any resource, has 2 a cost attendant with its usage. Unlike debt capital, 3 the cost of which can usually be determined from the 4 interest rate specified in the contract between the 5 lender and the utility, the rate of return investors 6 require from an investment in common equity is not so 7 readily ascertained. Nonetheless, common equity 8 investors still require a return on their investment, 9 with the "cost of equity" being the minimum rent that 10 must be paid for the use of their money. This cost of 11 equity typically serves as the starting point for 12 determining a fair rate of return on common equity. 13 Q. Does your analysis of a fair rate of return 14 on common equity for WWP assume that it is operating in 15 an unregulated market? 16 A. No. My analysis is based on the risks 17 currently faced by WWP as a utility operating under the 18 traditional regulatory compact. While greater 19 uncertainties surround currently regulated operations 20 because the ultimate impact of deregulation is unclear, 21 it is proper to reflect the risks associated with 22 restructuring the electric utility industry in current 23 authorized rates of return on common equity. 24 Q. What fundamental economic principle 25 underlies the cost of equity concept? 259 Avera, Di 30 WWP 1 A. The cost of equity concept is predicated on 2 the notion that investors are risk averse, and will 3 willingly bear additional risk only if they expect 4 compensation for their risk bearing. In capital markets 5 where relatively risk-free assets are available, such as 6 U.S. Treasury securities, investors can be induced to 7 hold more risky assets only if they are offered a 8 premium, or additional return, above the rate of return 9 on a risk-free asset. Since all assets compete with each 10 other 11 12 / 13 14 / 15 16 / 17 18 19 20 21 22 23 24 25 260 Avera, Di 30A WWP 1 for investors' funds, more risky assets must yield a 2 higher expected rate of return than less risky assets in 3 order for investors to be willing to hold them. 4 Given this risk-return tradeoff, the required rate 5 of return (k) from an asset (i) can be generally 6 expressed as: 7 ki = Rf + RPi 8 where: Rf= Risk-free rate of return; and Rpi= Risk premium required to hold 9 more risky asset i. 10 Thus, the required rate of return for a particular asset 11 at any point in time is a function of: 1) the yield on 12 risk-free assets, and 2) its relative risk, with 13 investors demanding correspondingly larger risk premiums 14 for assets bearing greater risk. 15 Q. Is there evidence that the risk-return 16 tradeoff principle actually operates in the capital 17 markets? 18 A. Yes. The risk-return tradeoff can be 19 readily documented in certain segments of the capital 20 markets where required rates of return can be directly 21 inferred from market data and generally accepted measures 22 of risk exist. For example, bond yields are reflective 23 of investors' expected rates of return, and bond ratings 24 are indicative of the risk of fixed income securities. 25 The observed yields on government securities and bonds of 261 Avera, Di 31 WWP 1 various rating categories demonstrate that the 2 risk-return tradeoff does, in fact, exist in the capital 3 markets. 4 To illustrate, average yields during October 1998 5 on selected U.S. government securities and on public 6 utility bonds of different ratings reported by Moody's 7 are shown in the following table. As evidenced there, as 8 risk increases (measured by progressively lower bond 9 ratings), the required rate of return (measured by 10 yields) rises accordingly. 11 12 / 13 14 / 15 16 / 17 18 19 20 21 22 23 24 25 262 Avera, Di 31A WWP 1 Also shown are the indicated risk premiums over long-term 2 government securities for the additional risk associated 3 with each bond rating category: 4 October 1998 Risk Premium Over Bond and Rating Yield Long-term Treasury 5 U.S. Treasury 6 5-Year 4.17% -- Long-term 5.21% -- 7 Public Utility 8 Aaa 6.64% 1.43% Aa 6.80% 1.59% 9 A 6.96% 1.75% Baa 7.13% 1.92% 10 11 Q. Does the risk-return tradeoff observed with 12 fixed income securities extend to common stocks and other 13 assets? 14 A. Documenting the risk-return tradeoff for 15 assets other than fixed income securities is complicated 16 by two factors. First, there is no standard measure of 17 risk applicable to all assets. Second, for most assets 18 (e.g., common stock), required rates of return cannot be 19 directly observed. Yet there is every reason to believe 20 that investors exhibit risk aversion in deciding whether 21 or not to hold common stocks and other assets, just as 22 when choosing among fixed income securities. 23 Accordingly, it is generally accepted that the 24 risk-return tradeoff evidenced with long-term debt 25 extends to all assets. 263 Avera, Di 32 WWP 1 The extension of the risk-return tradeoff from 2 assets with observable required rates of return (e.g., 3 bonds) to other assets is represented by the concept of a 4 "capital market line". In particular, competition 5 between securities and among investors in the capital 6 markets drives the prices of assets to equilibrium such 7 that the expected rate of return from each is 8 commensurate with its risk. Thus, 9 10 / 11 12 / 13 14 / 15 16 17 18 19 20 21 22 23 24 25 264 Avera, Di 32A WWP 1 the expected rate of return from any asset is a risk-free 2 rate of return plus a corresponding risk premium. This 3 concept of a capital market line is illustrated below. 4 The vertical axis represents required rates of return and 5 the horizontal axis indicates relative riskiness, with 6 the intercept of the capital market line being the 7 risk-free rate of return. 8 9 10 11 12 (Graph contained in hard copy of transcript.) 13 14 15 16 17 18 19 20 Q. Is this risk-return tradeoff limited to 21 differences between firms? 22 A. No. The risk-return tradeoff principle 23 applies not only to investments in different firms, but 24 also to different securities issued by the same firm. As 25 discussed earlier, the securities issued by a utility 265 Avera, Di 33 WWP 1 vary considerably in risk because they have different 2 characteristics and priorities. Long-term debt secured 3 by a mortgage on property is senior among all capital in 4 its claim on a utility's net revenues and is, therefore, 5 the least risky. Following first mortgage bonds are 6 other debt instruments also holding contractual claims on 7 the utility's net revenues, such as debentures. The last 8 investors in line 9 10 / 11 12 / 13 14 / 15 16 17 18 19 20 21 22 23 24 25 266 Avera, Di 33A WWP 1 are common shareholders. They only receive the net 2 revenues, if any, that remain after all other claimants 3 have been paid. As a result, the rate of return that 4 investors require from a utility's common stock, the most 5 junior and riskiest of its securities, must be 6 considerably higher than the yield offered by the 7 utility's senior, long-term debt. 8 Q. What does the above discussion imply with 9 respect to estimating the cost of equity? 10 A. Although the cost of equity cannot be 11 observed directly, it is a function of the returns 12 available from other investment alternatives and the 13 risks to which the equity capital is exposed. Because it 14 is unobservable, the cost of equity for a particular 15 utility must be estimated by analyzing information about 16 capital market conditions generally, assessing the 17 relative risks of the utility specifically, and employing 18 various quantitative methods that focus on investors' 19 required rates of return. These various quantitative 20 methods typically attempt to infer investors' required 21 rates of return from stock prices, by extrapolating 22 interest rates, or through an analysis of other financial 23 data. 24 Q. Did you rely on a single method to estimate 25 the cost of equity for WWP? 267 Avera, Di 34 WWP 1 A. No. Despite the theoretical appeal of or 2 precedent for using a particular method to estimate the 3 cost of equity, no single approach can be regarded as 4 wholly reliable. Therefore, I used both DCF and risk 5 premium methods to estimate the cost of equity for WWP. 6 Indeed, it is essential that estimates of investors' 7 required rate of return produced by one method be 8 compared with those produced by other methods, and that 9 all cost of equity estimates be required to pass 10 fundamental tests of 11 12 / 13 14 / 15 16 / 17 18 19 20 21 22 23 24 25 268 Avera, Di 34A WWP 1 reasonableness and economic logic. 2 B. Discounted Cash Flow Theory 3 Q. How are DCF models used to estimate the 4 cost of equity? 5 A. The use of DCF models to estimate the cost 6 of equity is essentially an attempt to replicate the 7 market valuation process which led to the price investors 8 are willing to pay for a share of a company's stock. It 9 is predicated on the assumption that investors evaluate 10 the risks and expected rates of return from all 11 securities in the capital markets. Given these expected 12 rates of return, the price of each share of stock is 13 adjusted by the market so that investors are adequately 14 compensated for the risks to which they are exposed. 15 Therefore, we can look to the market to determine what 16 investors feel a share of common stock is worth, and by 17 estimating the cash flows they expect to receive from the 18 stock in the way of future dividends and stock price, 19 their required rate of return can be mathematically 20 imputed. In other words, the cash flows that investors 21 expect from a stock are estimated, and given its current 22 market price, we can "back-into" the discount rate, or 23 cost of equity, investors presumably used in arriving at 24 that price. 25 Q. What market valuation process underlies DCF 269 Avera, Di 35 WWP 1 models? 2 A. DCF models are derived from a theory of 3 valuation which posits that the price of a share of 4 common stock is equal to the present value of the 5 expected cash flows (i.e., future dividends and stock 6 price) that will be received while holding the stock, 7 discounted at investors, required rate of return, or the 8 cost of equity. Notationally, the general form of the 9 DCF model is as follows: 10 11 D1 D2 Dt Pt P0 = + +... + + 12 (1+Ke) (1+Ke)2 (1+Ke) (1+Ke)t 13 14 / 15 16 / 17 18 / 19 20 21 22 23 24 25 270 Avera, Di 35A WWP 1 where: P0 Current price per share; Pt Future price per share in period t; 2 Dt Expected dividend per share in period t; Ke Cost of equity. 3 4 Q. Has this general form of the DCF model 5 customarily been used to estimate the cost of equity in 6 rate cases? 7 A. No. In an effort to reduce the number of 8 required estimates and computational difficulties, the 9 general form of the DCF model has been simplified to a 10 "constant growth" form. But converting the general form 11 of the DCF model to the constant growth DCF model 12 requires that a number of strict assumptions be made. 13 These include: 14 * A constant growth rate for both dividends and earnings; 15 * A stable dividend payout ratio; * The discount rate exceeds the growth rate; 16 * A constant growth rate for book value and price; * A constant earned rate of return on book value; 17 * No sales of stock at a price above or below book value; 18 * A constant price-earnings ratio; * A constant discount rate (i.e., no changes in risk 19 or interest rate levels and a flat yield curve); and 20 * All of the above extend to infinity. 21 Given these assumptions, the general form of the DCF 22 model can be reduced to the more manageable formula of: 23 D1 P0 = 24 ke - g 25 where: g = Investors' long-term constant growth expectations. 271 Avera, Di 36 WWP 1 The cost of equity (ke) can be isolated by rearranging 2 terms: 3 D1 ke = + g 4 P0 5 6 This constant growth form of the DCF model recognizes 7 that the rate of return to stockholders consists of two 8 parts: 1) dividend yield (Dj/P0), and 2) growth (g). In 9 other words, investors expect to receive 10 11 / 12 13 / 14 15 / 16 17 18 19 20 21 22 23 24 25 272 Avera, Di 36A WWP 1 a portion of their total return in the form of current 2 dividends and the remainder through price appreciation. 3 Q. To which firms did you apply the DCF model 4 to estimate the cost of equity for WWP? 5 A. Because conventional applications of the 6 DCF model to WWP are hindered by its recent dividend cut, 7 the DCF model was applied to proxy firms having publicly 8 traded shares. Publicly traded proxy firms for 9 WWP were selected from those electric utilities followed 10 by The Value Line Investment Survey (Value Line) and 11 included in S&P's Electric Utilities Group, which are 12 rated single-A by both Moody's and S&P. Included within 13 this group of electric utilities were companies 14 which also have gas distribution operations, as does WWP. 15 Excluded from the group were firms with nuclear 16 generation or that were currently involved in a merger. 17 These selection criteria resulted in the following group 18 of 8 electric utilities: 19 Black Hills Corp. Cleco Corp. 20 CINergy, Inc. Empire District Electric 21 New Century Energies Northwestern Corp. 22 Pacificorp Potomac Electric Power 23 24 C. Constant Growth DCF Model 25 Q. Are the assumptions underlying the constant 273 Avera, Di 37 WWP 1 growth form of the DCF model met in the real world? 2 A. No, none of the assumptions required to 3 convert the general form of the DCF model to the constant 4 growth form are ever strictly met in practice. In some 5 instances, where earnings are derived solely from stable 6 activities, and earnings, dividends, and book value track 7 fairly 8 9 / 10 11 / 12 13 / 14 15 16 17 18 19 20 21 22 23 24 25 274 Avera, Di 37A WWP 1 closely, the constant growth form of the DCF model may be 2 a reasonable working approximation of stock valuation. 3 However, in other cases, here the circumstances 4 surrounding the firm cause the required assumptions to be 5 severely violated, the constant growth DCF model may 6 produce widely divergent and meaningless results. This 7 is especially the case if the firm's earnings or 8 dividends are unstable, or if investors are expecting the 9 stock price to be affected by factors other than earnings 10 and dividends. 11 Q. How is the constant growth form of the DCF 12 model typically used to estimate the cost of equity? 13 A. The first step in implementing the constant 14 growth DCF model is to determine the expected dividend 15 yield (D1/P0) for the firm in question. This is usually 16 calculated based on an estimate of dividends to be paid 17 in the coming year divided by the current price of the 18 stock. The second, and more controversial, step is to 19 estimate investors' long-term growth expectations (g) for 20 the firm. Since book value, dividends, earnings, and 21 price are all assumed to move in lockstep in the constant 22 growth DCF model, estimates of expected growth are often 23 derived from historical rates of growth in these 24 variables under the presumption that investors expect 25 these rates of growth to continue into the future. 275 Avera, Di 38 WWP 1 Alternatively, a firm's internal growth can be estimated 2 based on the product of its earnings retention ratio (b) 3 and earned rate of return on equity (r). This growth 4 estimate may rely on either historical or projected data, 5 or both. A third approach is to rely on security 6 analysts' projections of growth in a firm's book value, 7 dividends, earnings, and stock price as proxies for 8 investors' expectations. The final step is to sum the 9 firm's dividend yield and estimated growth rate 10 11 / 12 13 / 14 15 / 16 17 18 19 20 21 22 23 24 25 276 Avera, Di 38A WWP 1 to arrive at an estimate of its cost of equity. 2 Q. How did you calculate the dividend yield 3 for each utility? 4 A. Value Line's estimate of dividends to be 5 paid by each of these electric utilities over the next 6 twelve months, obtained from the index to its September 7 11, October 9, and November 20, 1998 editions, served as 8 D1. This annual dividend was then divided by the recent 9 price reported by Value Line for each utility to 10 calculate the expected dividend yield. The expected 11 dividends, recent price, and resulting dividend yields 12 for each of the 8 single-A electric utilities are 13 displayed on Schedule WEA-4. As shown there, the 14 resulting dividend yield averaged 5.1 percent for the 15 proxy group. 16 Q. How did you measure growth in applying the 17 constant growth DCF model? 18 A. Given that the industry is becoming 19 increasingly competitive, diversified, and consolidated, 20 it is universally recognized that the future for electric 21 utilities will not be an extension of the past. Dividend 22 policy of electric utilities is also becoming 23 increasingly conservative as industry deregulation 24 becomes more pervasive. As noted in Value Line's March 25 13, 1998 edition: 277 Avera, Di 39 WWP 1 ... many companies that once deemed a payout ratio of 80% reasonable now seek to lower that ratio 2 because of uncertainties they face. (p. 160) 3 Because of this, historical growth rates and near-term 4 trends in dividends per share undoubtedly provide little 5 insight as to investors' current expectations for 6 electric utilities. 7 As a result, projected growth in earnings, which 8 ultimately support the dividends and capital appreciation 9 underlying share valuation, is likely to provide a more 10 plausible measure of "g" for use in the 11 12 / 13 14 / 15 16 / 17 18 19 20 21 22 23 24 25 278 Avera, Di 39A WWP 1 constant growth DCF model. Because internal, or 2 "sustainable", growth as been routinely examined in 3 regulatory proceedings and is consistent with the strict 4 theoretical assumptions of the constant growth DCF model, 5 "bxr" growth rates based on security analysts' near-term 6 projections were also developed for completeness. 7 Q. What did these measures imply with respect 8 to the "g" component of the constant growth DCF model? 9 A. One source of earnings per share (EPS) 10 growth projections widely referenced in the investment 11 literature and regulatory proceedings is I/B/E/S 12 International, Inc. (I/B/E/S), which reports average 13 5-year growth estimates compiled from over 2,300 14 financial analysts. The near-term EPS growth projections 15 reported by I/B/E/S for each of the 8 electric Utilities 16 in the group are displayed in Schedule WEA-4. As shown 17 there, security analysts are projecting average growth 18 rates for the group of single-A electric utilities in the 19 2.1 to 4.7 percent range, implying an average near-term 20 growth rate on the order of 3.7 percent. 21 Also presented on Schedule WEA-4 are the "bxr" 22 growth rates implied by Value Line's 2001-2003 23 projections of EPS, dividends per share (DPS), and net 24 book value per share (BV) for each of these electric 25 utilities. Internal growth for the group of 8 single-A 279 Avera, Di 40 WWP 1 electric utilities ranged from 2.4 to 6.7 percent, and 2 averaged 4.1 percent. 3 Q. Do the cost of equity estimates resulting 4 from these conventional applications of the constant 5 growth DCF model reasonably approximate what investors' 6 presently require from electric utility stocks? 7 A. No. As shown on Schedule WEA-4, applying 8 the constant growth 9 10 / 11 12 / 13 14 / 15 16 17 18 19 20 21 22 23 24 25 280 Avera, Di 40A WWP 1 DCF model using EPS and "bxr" growth rates based on 2 security analysts' projections implied an average cost of 3 equity for the group of 8 single-A electric utilities on 4 the order of 9 percent. All but two of the fifteen cost 5 of equity estimates were in the single-digits, with eight 6 falling below 9 percent. 7 It is inconceivable that investors' required rate 8 of return on equity, the most junior and risky of an 9 electric utility's securities, would not exceed the 7 10 percent average yield on single-A public utility bonds by 11 significantly more than 200 basis points. Given that the 12 uncertainties faced by electric utilities are almost 13 certainly greater now than at any time in recent history, 14 it is simply incongruous to assert that the cost of 15 equity has simultaneously declined to an unprecedented 16 low. Yet this is exactly the conclusion suggested by the 17 results of conventional applications of the constant 18 growth DCF model. 19 Q. Why does the constant growth DCF model fail 20 to reflect how investors are presently valuing electric 21 utility stocks? 22 A. The constant growth DCF model is predicated 23 on stable conditions; but, as described earlier, the 24 electric utility industry is in the midst of a major 25 upheaval. Competition is being increasingly promoted at 281 Avera, Di 41 WWP 1 the federal and state levels, and as a result of 2 deregulation and ensuing competition on both the supply 3 and demand sides of the industry, electric utilities' 4 traditional monopoly status is being eroded. This 5 transition of electric utilities to more competitive 6 markets is affecting investors' expectations in a variety 7 of ways, from the possibility of stagnant dividend growth 8 and earnings reductions in the near-term to prospects for 9 higher growth in the longer-term. 10 Meanwhile, many electric utility companies are 11 aggressively 12 13 / 14 15 / 16 17 / 18 19 20 21 22 23 24 25 282 Avera, Di 41A WWP 1 expanding and diversifying their operations. As noted in 2 S&P's Utilities & Perspectives (January 5, 1998): 3 With increasing industry competition, utilities are beginning to break out of old paradigms and 4 are seeking ways to differentiate themselves and create a competitive edge. (p.1) 5 6 Electric utilities have recently begun to merge and 7 acquire other domestic electric and/or gas utilities. 8 While some are pursuing investments in unrelated areas, 9 other major acquisitions have involved overseas electric 10 utility activities, which hold higher rate of return 11 prospects than those available from domestic operations. 12 For all of these reasons, it is unrealistic to 13 assume that electric utilities will grow prospectively at 14 a constant rate equal to near-term projections, or that 15 the future will be little more than an extension of the 16 past. The constant growth DCF model, as conventionally 17 applied, simply does not capture investors' long-term 18 expectations associated with increasing competition, 19 diversification, and consolidation in the electric 20 utility industry. As a result, traditional applications 21 of the constant growth DCF model produce cost of equity 22 estimates which are at direct odds with the realities 23 investors face in the capital markets. 24 Q. Is there evidence that investors expect 25 electric utilities to be significantly affected by the 283 Avera, Di 42 WWP 1 fundamental changes in the industry? 2 A. Yes. The investment literature is replete 3 with discussions of how the introduction of competition 4 is beginning, and will continue, to impact electric 5 utilities. As described earlier, these structural 6 changes so significantly affect investors' view of the 7 industry that S&P completely overhauled its bond rating 8 process for electric utilities. As Merrill Lynch 9 explained in its Electric Utilities Industry research 10 11 / 12 13 / 14 15 / 16 17 18 19 20 21 22 23 24 25 284 Avera, Di 42A WWP 1 report (June 24, 1996): 2 The electric utility industry is in a monumental transition state at the current time. The 3 transition is from a vertically-integrated, monopoly industry to one that we expect to be very 4 competitive and significantly restructured. We expect all utility customers to have competitive 5 choices in the next 5-10 years. We expect companies to realign and/or disaggregate their 6 businesses -- some may exit the generation business, others may exit the distribution 7 business -- as well as merge to create larger companies. ... The risk profile of the electric 8 utility industry is clearly reaching higher levels than it has experienced in the past and 9 will further increase ... (p. 3) 10 Similarly, Deregulation of the Electric Utility Industry: 11 An Overview (January 28, 1997), published by the 12 Association for Investment Management and Research 13 (AIMR), concluded: 14 Everything about the electric utility industry is undergoing a transformation. The basics of this 15 industry are no longer valid, which means new analytical tools are needed to understand and to 16 analyze electric utilities. Deregulation is redefining the environment in which the industry 17 operates and creating new challenges for industry participants. Industry restructuring is affecting 18 the valuation of electric utility securities, making investing in these securities more 19 challenging today than ever before. (p. 1) 20 And more recently, the February 19, 1998 edition of S&P's 21 Electric Utilities Industry Survey noted that: The electric power industry is in the midst of a 22 radical change. The monopolistic, tightly regulated utilities created under trust-busting 23 legislation more than 60 years ago are slowly being exposed to competition, particularly in the 24 generation and wholesale power markets. Technological advances and the increased 25 desire for customer choice are spurring the demand 285 Avera, Di 43 WWP 1 for new legislation. Whereas in the past electricity markets were strictly delineated by 2 geographic lines, utilities now have the freedom to cross into one another's territories. The 3 pressures of this growing competition will be exacerbated by the disparity in rates among 4 different regions. As aggressive, low-cost, well financed 5 players seek expansion under the new industry structure, the industry is likely to see 6 increasing consolidation and, ultimately, the emergence of some dominant powerhouse companies. 7 (p. 8) 8 9 / 10 11 / 12 13 / 14 15 16 17 18 19 20 21 22 23 24 25 286 Avera, Di 43A WWP 1 Q. Why are growth rates based on past 2 experience or near-term projections for electric 3 utilities unlikely to be indicative of what investors 4 expect over the longer-term? 5 A. As noted above, growth expectations for 6 electric utilities are clouded by the impact of 7 increasing competition in the industry. Nonetheless, it 8 is widely believed that near-term growth may tend to be 9 relatively modest as electric utilities prepare for a 10 more competitive market. But once the constraints of 11 regulation are relaxed and/or removed, electric utilities 12 are expected to enjoy growth rates more closely 13 paralleling those of competitive firms. Likewise, the 14 higher earnings prospects from electric utilities' 15 expanding involvement in non-regulated and overseas 16 utility activities are also likely to cause investors' 17 long-term growth expectations to be greater than 18 historical or near-term projected growth. This view of 19 differing near- and longer-term growth expectations for 20 electric utilities was corroborated by Merrill Lynch in 21 their Electric Utilities Industry report (January 8, 22 1997): 23 In the current environment, with excess capacity, an abundance of players, limited liquidity, and 24 the need to do strategic acquisitions, we believe that this could be a tough business to make money 25 for awhile. Ultimately, we believe there could 287 Avera, Di 44 WWP 1 be potential for strong returns, as the market reaches equilibrium and the number of players 2 dwindles. (p. 42) 3 Finally, as noted in S&P's Electric Utilities Industry 4 Survey (February 19, 1998): 5 Following in the footsteps of other deregulated industries -- such as airlines, 6 telecommunications, and banking -- investor-owned utilities have begun to consolidate. (p. 13) 7 8 But expectations of price appreciation that might be 9 realized in the event of a merger or acquisition are not 10 incorporated into the 11 12 / 13 14 / 15 16 / 17 18 19 20 21 22 23 24 25 288 Avera, Di 44A WWP 1 historical or projected growth estimates typically used 2 in the constant growth DCF model. 3 Q. How do differing near- and longer-term 4 expectations distort the cost of equity estimates 5 produced by conventional applications of the constant 6 growth DCF model? 7 A. Recall that the constant growth DCF model 8 assumes investors expect the same rate of growth to 9 prevail from now until infinity. Moreover, customary 10 applications of the constant growth DCF model simply 11 assume that historical experience or the near-term growth 12 projected by security analysts will continue into 13 perpetuity. However, if investors expect a utility's 14 growth in the longer-term to be higher, or lower, than 15 its growth in the near-term, then using just the 16 near-term growth rate will under- or over-state the 17 actual cost of equity. It is for this reason that the 18 constant growth DCF model is ill-suited to accommodating 19 significant differences between near-and longer-term 20 expectations. 21 Q. Have the failures of the constant growth 22 DCF model been recognized by regulators? 23 A. Yes. It is becoming increasingly evident 24 to regulatory agency staff members and commissioners that 25 conventional applications of the constant growth DCF 289 Avera, Di 45 WWP 1 model to utility stocks are not producing accurate 2 estimates of investors' required rates of return, with 3 increased reliance being placed on other methods to 4 estimate the cost of equity. For example, in a 1995 rate 5 case before the Public Utility Commission of Texas 6 involving Gulf States Utilities Company (Docket No. 7 12852), the Administrative Law Judge (ALJ) stated that: 8 9 / 10 11 / 12 13 / 14 15 16 17 18 19 20 21 22 23 24 25 290 Avera, Di 45A WWP 1 The constant DCF model should not be used in the changing environment of the electric utility 2 industry. No one disputes that the industry is facing a competitive future. With the advent of 3 competition, the factors that face the industry are uncertain. But it is not reasonable to assume 4 that the growth rates of a regulated monopoly will remain stagnate as that monopoly shifts to a 5 competitive market. Thus, the use of the constant DCF model is not reasonable. (p. 42) 6 7 The PUCT Staff has abandoned the constant growth DCF 8 model altogether and adopted a "multi-stage" DCF model in 9 which investors' longer-term growth expectations are 10 based in part on projected growth for the firms 11 comprising the market as a whole. 12 At the federal level, the FERC found in a July 7, 13 1994 order issued in the Ozark Gas Transmission System 14 case (Docket Nos. RP94-105-002 and RP94-105-003) that: 15 In summary, we have reversed the ALJ's sole reliance on five-year growth projections for the 16 DCF analysis, finding that the five-year projections are not of themselves incorrect, but 17 merely limited to too brief a time period to meet the requirements of the DCF-model. (p. 12) 18 19 The FERC made similar findings in a number of cases 20 involving natural gas pipelines, and the FERC Staff has 21 also abandoned the constant growth DCF model in testimony 22 filed in electric utility cases. In the Initial Decision 23 (December 22, 1994) in Docket No. EL94-24-000 involving 24 Allegheny Generating Company, the ALJ stated that: 25 The Commission clearly recognizes that there are 291 Avera, Di 46 WWP 1 ongoing changes in the electric industry which must be addressed. Id. at 31 & n.37 ("the electric 2 utility industry is continuing to evolve" and "[i}n recent months, the pace of change ... has 3 increased dramatically") (p. 4) 4 and then concluded that: 5 For all the above reasons (Commission precedent and the evidence in this record), I conclude that 6 the single-stage sustained growth method is inappropriate here. (p. 5) 7 8 9 Likewise, the constant growth DCF model was rejected in 10 four other FERC 11 12 / 13 14 / 15 16 / 17 18 19 20 21 22 23 24 25 292 Avera, Di 46A WWP 1 proceedings -- the November 17, 1995 Initial Decision in 2 the Southern Company Services, Inc. case (Docket No. 3 ER94-1348-000), the December 13, 1995 Initial Decision in 4 the Florida Power & Light Company case (Docket No. 5 ER93-465-000), the July 11, 1996 Initial Decision on 6 Threshold Issues in the System Energy Resources, Inc. 7 case (Docket No. ER95-1042000), and the November 27, 1996 8 Initial Decision in the Duke Power Company case (Docket 9 No, EL95-31-000). 10 Q. Considering these problems, did you rely on 11 the results of the constant growth DCF model to estimate 12 the cost of equity for WWP? 13 A. No. If investors' growth expectations for 14 electric utilities generally conform to the steady-state 15 assumptions underlying the constant growth DCF model, as 16 occurred in the past and may again be the case in the 17 future, this approach may provide a meaningful guide to 18 investors' required rate of return. But given the 19 structural changes occurring in the industry, investors 20 today clearly do not expect electric utilities' future 21 growth to be a repeat of the past, and near-term 22 projections are apt to understate the long-run growth 23 investors expect from the industry as the constraints of 24 regulation are relaxed. Therefore, conventional 25 applications of the constant growth DCF model to electric 293 Avera, Di 47 WWP 1 utilities cannot be expected to capture the rate of 2 return investors currently require from an equity 3 investment in this changing industry. Accordingly, I did 4 not rely on the results of the constant growth DCF model 5 to estimate the cost of equity for WWP, but instead 6 looked to DCF models which attempt to explicitly account 7 for the increasingly competitive nature of the electric 8 utility industry. 9 D. Non-Constant Growth DCF Model 10 Q. What alternatives are there to the constant 11 growth DCF model? 12 13 / 14 15 / 16 17 / 18 19 20 21 22 23 24 25 294 Avera, Di 47A WWP 1 A. There are basically two approaches to 2 overcome the infirmities currently associated with 3 conventional applications of the constant growth DCF 4 model to electric utilities. The first, often referred 5 to as a "two-stage" DCF model, attempts to develop a more 6 realistic growth rate by looking beyond just the 7 near-term and also capturing the longer-term growth 8 expectations investors have likely incorporated into the 9 current prices of common stocks. An average of the near- 10 and longer-term growth rates is then combined with the 11 electric utility's dividend yield to estimate the cost of 12 equity. 13 The second approach is a "multi-stage" DCF model, 14 which is essentially the general form of the DCF model 15 presented earlier. Application of a multi-stage DCF 16 model entails estimating the annual dividends investors 17 expect to receive from holding a share of stock and the 18 price at which they expect to sell it in the future. 19 These projected cash flows are then mathematically 20 equated to the current price of the stock to impute 21 investors' required rate of return. 22 Q. Which of these alternatives did you use to 23 develop DCF cost of equity estimates for WWP? 24 A. Although applications of the two-stage DCF 25 model may represent a considerable improvement over the 295 Avera, Di 48 WWP 1 constant growth DCF model, the two-stage model is limited 2 by the fact that it assumes only two discrete growth 3 rates. This hampers the two-stage DCF model's ability to 4 capture fully the impacts on investors' expectations of 5 increasing competition in the electric utility industry. 6 Therefore, I used a multi-stage DCF model, which 7 explicitly accounts for investors' expectations of 8 varying growth rates and payout ratios, and a transition 9 to deregulation for a segment of electric utilities' 10 present operations. This analysis is 11 12 / 13 14 / 15 16 / 17 18 19 20 21 22 23 24 25 296 Avera, Di 48A WWP 1 presented in Schedule WEA-5. 2 Q. Please describe generally the multi-stage 3 DCF model you used to estimate the cost of equity for 4 WWP. 5 A. The multi-stage DCF model I used 6 essentially assumes that investors expect electric 7 utilities' operations to be ultimately separated into 8 regulated (e.g., distribution and transmission) and 9 deregulated (e.g., generation) segments. It also assumes 10 that investors expect the deregulated segment to be fully 11 competitive in 10 years, with a transition period between 12 2002 and 2008. This model is based on the assumption 13 that investors expect each segment to grow at different 14 rates, with growth of the regulated segment reflecting 15 that of a conventional utility and growth in the 16 deregulated segment reflecting that of a competitive 17 firm. 18 Dividends expected by investors in the near-term 19 were based on company-specific dividend forecasts. 20 Expected dividends in the longer term were estimated 21 using a weighted average of the separate growth rates and 22 payout ratios for the regulated and deregulated segments, 23 as was a terminal stock price at the end of the 10-year 24 horizon. Finally, the cost of equity was imputed by 25 equating the projected dividend stream and future stock 297 Avera, Di 49 WWP 1 price to the firm's current stock price. 2 Q. Why did you assume that investors 3 distinguish between regulated and deregulated segments in 4 forming their long-term expectations for vertically 5 integrated electric utilities? 6 A. Virtually all of the discussions 7 surrounding the introduction of competition into the 8 electric utility industry envision the disaggregation of 9 formerly integrated operations and the deregulation 119 10 the generating function. For example, William I. Tilles, 11 then Vice 12 13 / 14 15 / 16 17 / 18 19 20 21 22 23 24 25 298 Avera, Di 49A WWP 1 President and Senior Analyst with Smith Barney, Inc. 2 noted in Interpreting Electric Utilities' Numbers and 3 Equity Valuation (January 28, 1997), published by AIMR: 4 The power-generation business is the business segment that will be most affected by the 5 transition to a competitive market. Of the three major business segments of a traditional utility, 6 power generation is the principal business segment that is likely to be operating in a completely 7 deregulated market in the near future. ... From a valuation perspective, generating assets should be 8 valued the same as deep cyclicals -- such as steel and pulp and paper -- with earnings multiples 9 possibly in the high single digits. (p. 35) 10 In a similar vein, S&P's September 1997 Industry 11 Commentary noted: 12 With monopoly structures and protective regulatory regimes becoming things of the past, competition 13 in parts of the [electric supply industries] will soon resemble other industries characterized by 14 commodity products, high fixed costs, long-lived assets, high capital investment, and low labor 15 content. (p. 3) 16 with S&P's October 1997 Global Sector Review concluding: 17 The typical unregulated generator is operating in a commodity business similar to refining, forest 18 products, or commodity chemicals. In addition, there is likely to be a shakeout in the generating 19 sector as competition really gets underway. (p. 17) 20 Investors clearly anticipate the disaggregation of 21 electric utilities along functional lines and recognize 22 that a deregulated market for electric generation implies 23 risks and prospects similar to those faced by other firms 24 in the competitive sector. 25 Q. Why did you assume that investors expect 299 Avera, Di 50 WWP 1 the deregulated segment to be fully competitive within 10 2 years? 3 A. Although the exact timetable remains 4 uncertain in most jurisdictions, there is every 5 indication that investors expect the deregulated portion 6 of the industry to be fully competitive within 7 approximately 10 years. As AIMR noted in Deregulation of 8 the Electric 9 10 / 11 12 / 13 14 / 15 16 17 18 19 20 21 22 23 24 25 300 Avera, Di 50A WWP 1 Utility Industry: An Overview (January 27, 1997): 2 ELECTRIC UTILITIES IN 2007 In 10 years, much of the deregulation and industry 3 restructuring that is only now beginning is likely to be fully accomplished. (p. 3) 4 Similarly, Value Line concluded in its January 9, 1998 5 report: 6 To date, only a few states have passed industry 7 restructuring legislation, but over the next two years, a significant number of jurisdictions will 8 likely enact new laws. State lawmakers and utility regulators are generally supporting bills 9 that phase in retail competition over a 5- to 10-year period. (p. 701) 10 11 Q. How were the annual dividend cash flows of 12 your multi-stage DCF model estimated? 13 A. Expected dividends in 1999 and 2002 were 14 equal to the company-specific projections published by 15 Value Line. DPS for the remaining years were calculated 16 as the product of annual EPS and the corresponding payout 17 ratio (P/0). 18 Q. Please describe in more detail the 19 assumptions underlying annual EPS for each electric 20 utility shown on Schedule WEA-5. 21 A. As with DPS, EPS for 1999 and 2002 were 22 equal to projections published by Value Line, with 23 estimates for 2000 and 2001 being interpolated. 24 Following the 2002 horizon of Value Line's forecast, EPS 25 were increased annually at the indicated growth rate. 301 Avera, Di 51 WWP 1 For 2008 and beyond, growth rates for the 2 regulated and deregulated segments were combined to 3 reflect investors' expectations for a conventional 4 utility and a competitive firm, respectively. 5 Specifically, growth for the regulated segment was 6 assumed equal to the average of the individual growth 7 rates reported by I/B/E/S for electric utilities, or 4.5 8 percent. For the deregulated segment, growth was based 9 on expectations for two competitive market benchmarks -- 10 the S&P 11 12 / 13 14 / 15 16 / 17 18 19 20 21 22 23 24 25 302 Avera, Di 51A WWP 1 500 Index and Value Line's Industrial Composite, which 2 consists of approximately 752 industrial, retail, and 3 transportation companies. Specifically, the S&P Earnings 4 Guide (October 1998) reported an average I/B/E/S growth 5 rate for the firms in the S&P 500 Index of 13 percent, 6 while Value Line's Selection & Opinion (August 21, 1998) 7 projected growth in EPS for the Industrial Composite of 8 10.0 percent. The resulting average growth rate of 11.5 9 percent for the deregulated segment was then weighted 10 equally with the 4.5 percent growth rate for conventional 11 utility to arrive at a composite growth rate for the 12 firm. 13 Finally, growth rates during the transition to 14 competition (2003 2007) were interpolated between the 15 company-specific growth rate implied by Value Line's 16 near-term EPS projections and the composite growth rate 17 for 2008 and beyond. 18 Q. Why did you weight the growth rates for the 19 regulated and deregulated segments equally? 20 A. This weighting reflects investors' 21 widespread belief that generating assets comprise at 22 least one-half of electric utilities total assets. For 23 example, Ellen Lapson, Senior Director of Fitch Investors 24 Service, L.P., noted in U.S. EIectric Industry: 25 Understanding the Basics (January 28, 1997) that: 303 Avera, Di 52 WWP 1 Generation predominates on the balance sheets, income statements, and cash flows of these 2 utilities and demands the largest share of management attention. Generation currently 3 represents about 59 percent of the book value of IOUs' assets and about 70 percent of the revenues. 4 The difference between the asset concentration and the revenue concentration reflects the fuel 5 intensity of the generation sector of the industry. (p. 7) 6 7 Likewise, in the article cited earlier, Smith Barney's 8 former Vice President and Senior Analyst, William I. 9 Tilles noted "the fact that 10 11 / 12 13 / 14 15 / 16 17 18 19 20 21 22 23 24 25 304 Avera, Di 52A WWP 1 about 50 percent of an electric company's assets are 2 invested in the power generation business segment" 3 (p. 34). 4 Q. What was the basis for the annual payout 5 ratios (P/0) shown on Schedule WEA-5? 6 A. Once again, Value Line's near-term 7 projections served as the source for values in 1999 and 8 2002, with P/O in 2000 and 2001 being interpolated. 9 In response to the transition to competition, 10 however, investors generally expect electric utilities to 11 adopt more conservative financial policies, including 12 dividend payout ratios more comparable to other firms in 13 the unregulated sector. As S&P observed in a September 14 26, 1996 Industry Survey: 15 While earnings are on the rise, dividends aren't keeping pace. The reason for this is that 16 more utilities are seeking to increase their cash on hand, so that they can expand or otherwise cope 17 in the face of heightened competition. Evidence of this trend is seen in the 18 dividend payout ratio, which measures the percentage of earnings paid out, or the annual 19 dividend per share. Industry-wide this ratio (after extraordinary items and excluding 20 nonrecurring items) decreased from 80% for the 12 months ending March 31, 1995, to 74.7 percent for 21 the comparable 12 months ending March 31, 1996. (pp. 5-6) 22 23 Accordingly, after 2002 P/O for each utility was 24 interpolated between Value Line's near-term, 25 company-specific forecast and an average ratio reflecting 305 Avera, Di 53 WWP 1 expected dividend policies for electric utilities' 2 regulated and deregulated segments. Assigning equal 3 weight to representative payout ratios for a conventional 4 utility and a competitive firm of 80 and 40 percent, 5 respectively, resulted in a composite P/O of 60 percent 6 Q. Is this view of differing expectations for 7 regulated and deregulated segments of electric utilities' 8 operations corroborated by 9 10 / 11 12 / 13 14 / 15 16 17 18 19 20 21 22 23 24 25 306 Avera, Di 53A WWP 1 participants in the capital markets? 2 A. Yes. In an article entitled Fearless 3 Forecast: Electric Utilities in 2007 published by AIMR 4 (January 28, 1997), Smith Barney, Inc. Senior Industry 5 Advisor Leonard S. Hyman noted that: 6 Dividend payout as a percentage of reported net income for IOU's is almost twice as much as the 7 S&P 400 industrials -- 78 percent versus 44 percent. (p. 65) 8 and concluded: 9 More than half of the industry's assets are tied 10 up in generation, a business rapidly turning competitive. Whether utilities retain generating 11 assets or not, they own them now. They require financial policies that meld regulated and 12 competitive elements. (p. 65) 13 Mr. Hyman went on to develop a composite payout ratio of 14 59 percent assuming that: 15 ... 55 percent of the industry (generation) must bring its practices in line with competitive 16 industries and that the balance of the utility industry keeps utility-type ratios. (p. 65) 17 18 Q. How did you calculate a future stock price 19 for 2008? 20 A. Investors are likely to expect the 21 regulated and deregulated segments of the electric 22 utility industry to reach a relatively steady 23 state once the transition to competition is completed. 24 Accordingly, an expected stock price at the end of 2008 25 was estimated using the constant growth DCF valuation 307 Avera, Di 54 WWP 1 model presented earlier, which capitalizes an 2 expected dividend (D2009) by the difference between the 3 cost of equity and expected long-term growth (k-g). This 4 constant growth form of the DCF model is applied only 5 after considering investors' changing expectations as 6 competition is introduced in the electric utility 7 industry. 8 9 / 10 11 / 12 13 / 14 15 16 17 18 19 20 21 22 23 24 25 308 Avera, Di 54A WWP 1 Q. What cost of equity is implied using the 2 multi-stage DCF model? 3 A. The costs of equity implied for each firm 4 in the single-A group are displayed on Schedule WEA-5. 5 As summarized on page 2 of that schedule, the indicated 6 costs of equity for the group of 8 electric utilities 7 ranged from 11.1 to 11.8 percent, and averaged 11.5 8 percent. 9 Q. Is the model developed above the only 10 possible application of the multi-stage DCF model to 11 estimate the cost of equity for WWP? 12 A. No. Because investors' expectations are 13 inherently unobservable, assumptions must be made in 14 order to implement any DCF model. As a result, the 15 multi-stage DCF model is necessarily based on number of 16 assumptions regarding investors' expectations and 17 beliefs, and changing any of them will impact the 18 estimated cost of equity. For example, shortening the 19 competitive transition period would result in higher cost 20 of equity estimates, as would increasing the weighting of 21 the deregulated segment. Of course, the converse would 22 have the opposite effect. 23 Given the structural changes that the electric 24 utility industry is undergoing, however, the assumptions 25 made above to implement the multistage DCF model 309 Avera, Di 55 WWP 1 reasonably reflect investors' current expectations. And, 2 while the greater number of required inputs increases the 3 apparent complexity of multi-stage DCF models, 4 considering the deregulation and increasing competition 5 faced by electric utilities, this approach is certainly 6 more plausible than the steady-state assumptions 7 underlying conventional applications of the constant 8 growth DCF model. 9 Q. Do you have any observations about the use 10 of DCF models to estimate the cost of equity? 11 12 / 13 14 / 15 16 / 17 18 19 20 21 22 23 24 25 310 Avera, Di 55A WWP 1 A. Yes. DCF models, and the constant growth 2 form in particular, have been one of the mainstays of 3 regulation for over 15 years. But because the cost of 4 equity is inherently unobservable, no single method 5 should be considered a wholly reliable guide to 6 investors' required rate of return. As the Federal 7 Communications Commission recognized in its 1995 Report 8 and Order evaluating the methodology used to prescribe 9 rates of return for telephone companies (CC Docket 10 No. 92-133): 11 Equity prices are established in highly volatile and uncertain capital markets. ... Different 12 forecasting methodologies compete with each other for eminence, only to be superceded by other 13 methodologies as conditions change. ... In these circumstances, we should not restrict ourselves to 14 one methodology, or even a series of methodologies, that would be applied mechanically. 15 Instead, we conclude that we should adopt a more accommodating and flexible position (pp. 42-43) 16 17 Regardless of how carefully performed or theoretically 18 consistent a particular DCF application may be, 19 consideration should also be given to the results of 20 other methods. 21 E. Risk Premium Analyses 22 Q. What other analyses did you conduct to 23 estimate the cost of equity? 24 A. The cost of equity was also evaluated using 25 various risk premium methods. The findings of leading 311 Avera, Di 56 WWP 1 studies of equity risk premiums for utilities reported in 2 the academic and trade literature were used as the basis 3 for estimating equity risk premiums. These studies 4 employed numerous approaches to estimate equity risk 5 premiums, and encompassed a variety of time periods and 6 sample groups of utilities. Because of this diversity, 7 certain adjustments were required to adapt the findings 8 of the studies to present capital market conditions and 9 reflect the relative risks of the single-A rated electric 10 11 / 12 13 / 14 15 / 16 17 18 19 20 21 22 23 24 25 312 Avera, Di 56A WWP 1 utility group. The details of my risk premium analyses 2 are contained in Appendix B, with the results being 3 summarized below. 4 Q. Briefly describe the risk premium method. 5 A. The risk premium method to estimate 6 investors' required rate of return extends the 7 risk-return tradeoff observed with bonds to common 8 stocks. The cost of equity is estimated by determining 9 the additional return investors require to forego the 10 relative safety of bonds and to bear the greater risks 11 associated with common stock, and then adding this equity 12 risk premium to the current yield on bonds. Like the DCF 13 model, risk premium analyses are capital market oriented, 14 but unlike DCF methods where the cost of equity is 15 indirectly imputed, risk premium methods estimate 16 investors' required rate of return directly by adding an 17 equity risk premium to observable bond yields. 18 Q. How is the risk premium method implemented? 19 A. The actual measurement of equity risk 20 premiums is complicated by the inherently unobservable 21 nature of the cost of equity. In other words, like the 22 cost of equity itself and the growth component of the DCF 23 model, equity risk premiums cannot be calculated 24 precisely. Therefore, equity risk premiums must be 25 estimated, with studies in the academic and trade 313 Avera, Di 57 WWP 1 literature typically relying on three general approaches 2 to obtain observable proxies for equity risk premiums: 3 1) expectational estimates of the cost of equity, 2) 4 surveys, and 3) realized rates of return. 5 Q. Is there any risk premium behavior which 6 needs to be considered when implementing the risk premium 7 method? 8 A. Yes. There is considerable evidence that 9 the magnitude of equity risk premiums is not constant, 10 and that equity risk premiums tend 11 12 / 13 14 / 15 16 / 17 18 19 20 21 22 23 24 25 314 Avera, Di 57A WWP 1 to move inversely with interest rates. Indeed, this 2 inverse relationship is evident in most of the studies 3 discussed in Appendix B. 4 Q. What did your various analyses imply with 5 respect to the current equity risk premium for the 6 electric utility group? 7 A. As discussed in detail in Appendix B, the 8 various methods I examined indicated an equity risk 9 premium over single-A public utility bond yields in the 10 3.75 to 5.25 percent range. Adding this equity risk 11 premium range to the October average 1998 yield on 12 single-A public utility bonds of approximately 7 percent 13 implied a current cost of equity for a single-A rated 14 electric utility on the order of 10.75 to 12.25 percent. 15 F. Summary of Findings 16 Q. How did you go about estimating the rate of 17 return investors presently require from WWP? 18 A. Given WWP's recent dividend restructuring 19 plan, my analyses of the cost of equity for WWP focused 20 on estimates of investors' required rate of return for a 21 proxy group of other electric utilities that, like WWP, 22 are rated single-A. Two general approaches were used to 23 analyze the cost of equity for WWP -- the DCF method and 24 the risk premium method. Both methods are capital market 25 oriented and widely used to estimate the rate of return 315 Avera, Di 58 WWP 1 investors require from an investment in the common stock 2 of public utilities. 3 Q. What were the results of your DCF analyses? 4 A. Conventional applications of the constant 5 growth DCF model were initially used in an attempt to 6 estimate the cost of equity for WWP. But because the 7 electric utility industry is in the midst of a major 8 structural change, it is unreasonable to assume, as the 9 constant 10 11 / 12 13 / 14 15 / 16 17 18 19 20 21 22 23 24 25 316 Avera, Di 58A WWP 1 growth DCF model does, that the future for electric 2 utilities will be little more than an extension of the 3 past, or that investors presently expect electric 4 utilities to grow into perpetuity at a constant rate 5 equal to near-term projections. 6 The inability of traditional applications of the 7 constant growth DCF model to capture investors' current 8 expectations for the electric utility industry was 9 readily apparent. Of the 15 cost of equity estimates 10 produced by this approach for the proxy group, all but 11 two were in the single-digits. My constant growth DCF 12 analysis indicated cost of equity on the order of 9 13 percent, barely 200 basis points above the average yield 14 on single-A rated first mortgage bonds. Considering the 15 heightened uncertainties associated with the transition 16 to competition, it is simply not credible to suggest that 17 investors do not require a significantly larger premium 18 to compensate for the greater risk of holding common 19 equity. 20 To overcome these infirmities, I relied on the 21 non-constant growth form of the DCF model to estimate the 22 cost of equity for WWP. Application of a multi-stage DCF 23 model, which explicitly accounts for investors' 24 expectations of varying growth rates and a changing mix 25 of regulated and deregulated activities, produced cost of 317 Avera, Di 59 WWP 1 equity estimates for the group of single-A rated electric 2 utilities ranging between 11.1 and 11.8 percent, and 3 averaging 11.5 percent. 4 Q. What were the results of your risk premium 5 analyses? 6 A. Risk premium analyses based on alternative 7 approaches and studies, updated to reflect present 8 capital market conditions, suggested a cost of equity 9 range for the single-A electric utility group on the 10 order of 10.75 to 12.25 percent. 11 12 / 13 14 / 15 16 / 17 18 19 20 21 22 23 24 25 318 Avera, Di 59A WWP 1 Q. What is your conclusion as to the cost of 2 equity? 3 A. The analyses described above implied that 4 the cost of equity for the group of single-A electric 5 utilities is in the range of approximately 11.0 to 12.0 6 percent. This range encompasses the 11.1 to 11.8 percent 7 cost of equity range indicated by my multi-stage DCF 8 analyses and overlaps all but the bottom- and top-most 9 portions of the 10.75 to 12.25 percent range implied by 10 the risk premium approach. 11 G. Other Factors 12 Q. Are there any other costs that should be 13 considered in setting an allowed rate of return on common 14 equity? 15 A. Yes. The common equity used to finance 16 utility assets is provided from either the sale of stock 17 in the capital markets or from retained earnings not paid 18 out as dividends. When equity is raised through the sale 19 of stock, there are costs associated with "floating" the 20 new equity securities. These flotation costs include 21 services such as legal, accounting, and printing, as well 22 as the fees and discounts paid to compensate brokers for 23 selling the stock to the public. Also, some argue that 24 the "market pressure" from the additional supply of 25 common stock and other market factors may further reduce 319 Avera, Di 60 WWP 1 the amount of funds a utility nets when it issues common 2 equity. 3 Q. Is there an established mechanism for a 4 utility to recognize equity flotation costs? 5 A. No. While debt flotation costs are 6 recorded on the books of the utility and amortized over 7 the life of the issue, serving to increase the effective 8 cost of debt capital, there is no similar accounting 9 treatment to ensure that equity flotation costs are 10 recorded and ultimately recognized. Alternatively, no 11 rate of return is 12 13 / 14 15 / 16 17 / 18 19 20 21 22 23 24 25 320 Avera, Di 60A WWP 1 authorized on flotation costs necessarily incurred to 2 obtain a portion of the equity capital used to finance 3 plant. In other words, equity flotation costs are not 4 included in a utility's rate base since neither that 5 portion of the gross proceeds from the sale of common 6 stock used pay flotation costs is available to invest in 7 plant and equipment, nor are flotation costs capitalized 8 as an intangible asset. Even though there is no 9 accounting convention to accumulate the flotation costs 10 associated with past equity issues, flotation costs are a 11 necessary expense of obtaining equity capital. Unless 12 some provision is made to recognize these issuance costs, 13 a utility's revenue requirements will not fully reflect 14 all of the costs incurred for the use of investors' 15 funds. 16 Q. How can equity flotation costs be 17 recognized in revenue requirements? 18 A. As indicated above, there is no direct 19 mechanism to recognize flotation costs necessarily 20 incurred in connection with the issuance of common stock 21 as there is with debt. Therefore, flotation costs must 22 be accounted-for indirectly, with an upward adjustment to 23 the cost of equity being the most logical mechanism to 24 reflect these costs. Indeed this is essentially how 25 flotation costs incurred in connection with the issuance 321 Avera, Di 61 WWP 1 of preferred stock are generally recognized, since the 2 cost of preferred stock is typically calculated by 3 dividing annual preferred dividend requirements by the 4 net proceeds from the sale of the stock issue. By using 5 net proceeds instead of face value as the denominator, 6 flotation costs are recognized in the resulting cost of 7 preferred stock. 8 9 / 10 11 / 12 13 / 14 15 16 17 18 19 20 21 22 23 24 25 322 Avera, Di 61A WWP 1 Q. What is the magnitude of the adjustment to 2 the "bare bones" cost of equity to account for flotation 3 costs? 4 A. There are any number of ways in which a 5 flotation cost adjustment can be calculated, with the 6 adjustment ranging from just a few basis points to more 7 than a full percent. For example, relating past 8 flotation costs to total book common equity normally 9 results in a nominal flotation cost adjustment of a few 10 basis points, while applying an average flotation cost 11 expense percentage (i.e., 3 to 5 percent) to a utility's 12 dividend yield, or its cost of equity, usually result in 13 flotation cost adjustment between 15 and 50 basis points. 14 Q. Is there any other factor that should be 15 considered in setting a fair rate of return on common 16 equity for WWP? 17 A. Yes. The 11 to 12 percent bare bones cost 18 of equity range developed above was based on analyses of 19 investors' required rate of return from a group of 8 20 publicly traded, single-A rated electric utilities. As 21 discussed in Section III, while generally comparable to 22 WWP, the average capitalization for the proxy group 23 contains a significantly greater proportion of common 24 equity than WWP's requested capital structure. A lower 25 common equity ratio translates into increased financial 323 Avera, Di 62 WWP 1 risk and, as a result, common shareholders require a 2 correspondingly higher rate of return to compensate them 3 for their risk bearing. Accordingly, if customers are to 4 benefit from the lower costs associated with WWP's 5 requested capital structure, then it is only proper that 6 they bear any higher costs required to obtain this 7 benefit. 8 Q. How does the PCA impact the relative risks 9 faced by WWP? 10 A. By providing a more responsive alternative 11 to recover higher power supply costs during low water 12 years, the PCA acts to moderate the 13 14 / 15 16 / 17 18 / 19 20 21 22 23 24 25 324 Avera, Di 62A WWP 1 impact of extreme water conditions on WWP's earnings. 2 However, this reduction in volatility is a double-edged 3 sword, since WWP foregoes greater earnings in years of 4 abundant streamflow. Thus, while the PCA may reduce the 5 risk associated with year-to-year extremes in water 6 conditions, WWP largely gives up the offsetting gains 7 previously realized in wet years. 8 It is important to note that the PCA does not 9 eliminate the underlying source of power production and 10 supply cost volatility. Whereas WWP's capacity is 11 dominated by hydroelectric facilities, most other 12 electric utilities have little, if any, hydro generation 13 on their systems. As a result, they maintain a fairly 14 stable generation mix, with fluctuations in power 15 production and supply costs being dampened by long-term 16 contracts. Thus, the PCA largely addresses a risk factor 17 not faced by most other electric utilities. Moreover, 18 the PCA does not improve WWP's competitive position since 19 the underlying generating cost structure remains 20 unchanged. Finally, because the vast majority of other 21 regulatory jurisdictions already allow the use of some 22 form of adjustment clause to recover power production and 23 supply costs, the PCA only brings WWP more into line with 24 the rest of the electric utility industry. 25 H. Conclusion 325 Avera, Di 63 WWP 1 Q. What then is your conclusion as to the fair 2 rate of return on equity for WWP? 3 A. As indicated earlier, based on the various 4 capital market oriented analyses described in my 5 testimony, I concluded that the "bare bones" cost of 6 equity for the group of other single-A electric utilities 7 is presently in the range of 11.0 to 12.0 percent. This 8 "bare bones" 9 10 / 11 12 / 13 14 / 15 16 17 18 19 20 21 22 23 24 25 326 Avera, Di 63A WWP 1 cost of equity, however, does not recognize flotation 2 costs incurred in connection with past and future sales 3 of common stock, nor does it account for the greater 4 financial risk implied by WWP's requested capital 5 structure. Accordingly, I added a minimal adjustment to 6 this range to arrive at a fair rate of return on common 7 equity range for WWP of 11.25 to 12.25 percent. I 8 recommend that WWP be authorized a rate of return on 9 common equity at the midpoint of this range, or 11.75 10 percent. 11 Q. Does this fair rate of return provide for 12 or recognize any incentive return for other factors? 13 A. No, it does not. My 11.75 percent 14 recommended fair rate of return on equity does not 15 explicitly incorporate any allowance for exemplary 16 performance or efficient and economic management, as 17 discussed in the testimony of Mr. Dukich. An incentive 18 return for such factors should be added to my fair rate 19 of return on equity for WWP. 20 21 22 23 24 25 327 Avera, Di 64 WWP 1 V. OVERALL RATE OF RETURN 2 Q What overall rate of return do you 3 recommend be applied to the original cost invested 4 capital of WWP? 5 A. I recommend that WWP be authorized an 6 overall rate of return on rate base of 9.446 percent. As 7 developed on Schedule WEA-6, this overall rate of return 8 is based on a capital structure consisting of 9 approximately 48.0 percent long-term debt, 4.0 percent 10 short-term debt, 10.6 percent preferred securities, and 11 37.4 percent common equity. This capital structure was 12 combined with the average costs of long-term debt, 13 short-term debt, and preferred securities discussed in 14 Section III of my testimony. As discussed in the 15 testimony of Mr. Dukich, the 12 percent component cost of 16 common equity is the sum of my 11.75 percent recommended 17 fair rate of return and a 25 basis-point premium to 18 reward WWP's exemplary performance and efficiency. 19 Q. Does this conclude your direct testimony in 20 this case? 21 A. Yes, it does. 22 23 24 25 328 Avera, Di 65 WWP 1 I. INTRODUCTION 2 Q. Please state your name and business 3 address. 4 A. William E. Avera, 3907 Red River, Austin, 5 Texas, 78751. 6 Q. Are you the same William E. Avera who 7 previously filed direct testimony in this case? 8 A. Yes, I am. 9 Q. What is the purpose of your rebuttal 10 testimony in this case? 11 A. My purpose here is to respond to the 12 testimony of Ms. Terri Carlock on behalf of the Staff of 13 the Idaho Public Utilities Commission (IPUC), and 14 Mr. Dennis E. Peseau on behalf of Potlatch Corporation 15 concerning the cost of equity for Avista Corp. (Avista), 16 formerly The Washington Water Power Company (WWP). 17 Q. Are you sponsoring any exhibits to be 18 introduced in this proceeding? 19 A. Yes, I am. My exhibit consists of four 20 schedules and has been marked for identification as 21 Exhibit No. 22. 22 Q. Please summarize your findings with respect 23 to the rate of return on equity recommendations of 24 Ms. Carlock and Mr. Peseau. 25 A. To her credit, Ms. Carlock recognized that 329 Avera, Di-Reb 1 Avista 1 the constant growth DCF model is currently ill suited to 2 estimating the cost of equity for Avista. Additionally, 3 she properly incorporated an adjustment for common stock 4 flotation costs and recommended a 25 basis point adder to 5 reward the accomplishments of Avista's management. 6 Nevertheless, as discussed in greater detail in my 7 testimony, Ms. Carlock's recommendation understated the 8 return investors require from Avista because: 9 * Comparable earnings analyses do not focus on the rates of return investors require in the capital 10 markets. Instead, they are inherently accounting- 11 12 13 / 14 15 / 16 17 / 18 19 20 21 22 23 24 25 330 Avera, Di-Reb 1A Avista 1 based and typically rely on historical information. Ms. Carlock placed undue emphasis on 2 operating results for a single year, 1998, which understated electric utilities' earned rates of 3 return. More comprehensive analyses of her comparable earnings data support a cost of equity 4 in the 11 to 12 percent range; 5 * Rather than applying a two-stage DCF model, as she suggested, the growth rates used in Ms. Carlock's 6 DCF analysis were an average of security analysts' projections for a single 3-5 year period. As a 7 result, her DCF model fails to match current share prices with the long-term growth expectations used 8 to arrive at these market values. Application of Ms. Carlock's DCF model using growth rates more 9 indicative of investors' expectations results in cost of equity estimates ranging from 10 approximately 10.9 to 12.7 percent; 11 * While Ms. Carlock concluded that Avista's risks were lower than for other utilities, the risk 12 measures she examined indicate that Avista is comparable to the firms she used to estimate the 13 cost of equity. Similarly, while the Power Cost Adjustment (PCA) moderates the impact of extreme 14 weather conditions, this only brings Avista in line with the rest of the industry; 15 * Ms. Carlock ignored the significantly greater 16 financial risk implied by Avista's regulatory capital structure, which contains much higher debt 17 levels than are maintained by the firms she used to estimate the cost of equity. As a result, her 18 cost of equity estimates do not incorporate the additional return investors require to bear the 19 greater financial risk associated with Avista's regulatory capital structure; and, 20 * Ms. Carlock's overall rate of return 21 recommendation is insufficient to maintain Avista's financial integrity, and thus fails to 22 meet the regulatory and economic standards established by judicial precedent. 23 With respect to Mr. Peseau's recommendations I concluded 24 that: 25 331 Avera, Di-Reb 2 Avista 1 * Rather than basing his conclusions on independent analyses of the cost of equity for Avista, 2 Mr. Peseau simply adjusted a single allowed rate of return authorized in 1986 based on changes in 3 Treasury bond yields; 4 * Because the cost of equity does not change in lockstep with interest rates, Mr. Peseau's simple 5 minded approach drastically understated investors' required rate of return from Avista; and, 6 Mr. Peseau entirely ignored the fact that investors' perceptions of the investment risks 7 associated with electric utilities have changed significantly since 1986. 8 9 10 / 11 12 / 13 14 / 15 16 17 18 19 20 21 22 23 24 25 332 Avera, Di-Reb 2A Avista 1 II. TERRI CARLOCK 2 Q. What cost of equity did Ms. Carlock 3 recommend for Avista? 4 A. Ms. Carlock recommended a cost of equity 5 range for Avista of between 10.25 and 11.25 percent. 6 Ms. Carlock's range was based on comparable earnings 7 analyses, which looked at rates of return on book equity 8 earned by industrial firms and electric and gas 9 utilities. In addition, she applied the discounted cash 10 flow (DCF) method to three groups of electric utilities 11 selected from the Electric Utility-West, Electric 12 Utility-Central, and Electric Utility-East industry 13 groups followed by The Value Line Investment Survey 14 (Value Line). Incorporated into her DCF analyses was an 15 adjustment for common stock flotation costs of 16 approximately 25 basis points. Ms. Carlock also 17 recommended adding 25 basis points to the 10.75 percent 18 midpoint of her cost of equity range to recognize 19 exemplary management performance, resulting in her 20 recommended fair rate of return on equity for Avista of 21 11.0 percent. 22 Q. Please describe Ms. Carlock's comparable 23 earnings analyses. 24 A. Ms. Carlock examined the rates of return on 25 book equity earned by industrial firms, as well as by 333 Avera, Di-Reb 3 Avista 1 electric and gas utilities. For industrials, she 2 presented average earned rates of return on book equity 3 over the years 1988 through 1998 for various industry 4 groups, as reported in Business Week's Corporate 5 Scoreboard. These included earned rates of return on 6 book equity calculated for the 12-months ending in each 7 quarter over the 11-year period, averaged over 8 alternative 5- and 3-year periods, and based on 3-year 9 moving averages. 10 Ms. Carlock also reviewed earned rates of return 11 on book equity for Moody's electric and natural gas 12 distribution utilities since 1970, including averages for 13 recent 10-, 5-, and 3-year periods. Based on her review 14 of the rates of return on book equity earned by 15 industrial 16 17 / 18 19 / 20 21 / 22 23 24 25 334 Avera, Di-Reb 3A Avista 1 and regulated firms, as well as a variety of other data 2 (e.g., interest rates, stock market price indices, and 3 Avista's relative risks), Ms. Carlock concluded that: 4 Using the comparable earnings approach, my estimate of the current cost of equity capital for 5 Avista is in the range of 10.5%-11.5%. (p. 15) 6 Q. What problems are associated with using 7 comparable earnings analyses to estimate investors' 8 required rate of return on equity? 9 A. In contrast to the capital market-oriented 10 DCF and risk premium methods, which estimate the rate of 11 return investors currently require for the use of their 12 money, comparable earnings analyses such as those 13 conducted by Ms. Carlock only tell us what rates of 14 return companies earned in the past. Because such 15 comparable earnings are inherently accounting-based and 16 backward looking, they are not particularly helpful in 17 identifying either the compensation investors currently 18 require for the use of their money or the rates of return 19 available from alternative investments in the capital 20 markets. Earned rates of return based on accounting 21 measures can be unduly influenced by changes in 22 accounting conventions or distorted by transitory 23 conditions. Meanwhile, earned rates of return on equity 24 based on historical information are not necessarily 25 indicative of investors' long-run perceptions of risk and 335 Avera, Di-Reb 4 Avista 1 expectations for returns going forward. This is 2 especially the case for the electric power industry, 3 where investors anticipate a transition to competitive 4 markets of various types. 5 Q. Are near-term projections of earned rates 6 of return on book equity for electric utilities likely to 7 provide a more meaningful guide to the rate of return 8 required by investors? 9 A. No. Ms. Carlock referenced the earned 10 rates of return projected by Value Line for the next 3-5 11 years, which ranged from 7.5 to 8.3 percent for the 12 electric utilities in her comparable groups. But, as 13 discussed in Section IV of my direct testimony, the 14 transition to competition in the electric power industry 15 is widely expected to constrain utility earnings 16 17 / 18 19 / 20 21 / 22 23 24 25 336 Avera, Di-Reb 4A Avista 1 during the period typically covered by security analysts' 2 near-term projections. Moreover, the returns on book 3 equity that utilities may be expected to earn tell us 4 little about the rates of return investors require in the 5 marketplace. Indeed, with single-A rated public utility 6 bonds yielding an average of approximately 7.3 percent in 7 March 1999, Value Line's near-term projected earned rates 8 of return for electric utilities equate to an equity risk 9 premium ranging from 20 to 100 basis points. Clearly, 10 investors require a substantially higher premium for 11 holding residual common stock, the most junior of 12 securities. 13 Q. Was Ms. Carlock's heavy reliance on recent 14 historical information warranted? 15 A. No. Despite presenting 11 years of 16 historical data from Business Week and 29 years of 17 information for Moody's electric and gas distribution 18 utilities, Ms. Carlock apparently placed little, if any, 19 weight on the rates of return earned on book equity 20 before 1995. Nine of the fourteen earned rates of return 21 Ms. Carlock cited as support for her comparable earnings 22 cost of equity range were based solely on operating 23 results for 1998, which in many cases were adversely 24 affected by abnormal weather conditions. As Value Line 25 (July 10, 1998) noted: 337 Avera, Di-Reb 5 Avista 1 Unusual weather has had a notable impact on operating results so far this year... During the 2 first quarter of this year, temperatures were milder than normal, restraining revenue and 3 earnings advances... Cool and wet weather prevailed in April and May, further pressuring 4 profits. (p. 701) 5 In turn, moderate temperatures depressed the 1998 earned 6 rates of return for the utilities included in Ms. 7 Carlock's benchmark groups. For example, with respect to 8 Nevada Power Company Value Line (August 21, 1998) 9 observed that: 10 Mild weather hurt second-quarter earnings. Revenues fell below 1997's level and earnings 11 declined by 50%. In all, we expect 1998 share net to remain flat. (p. 1734) 12 13 Meanwhile, an unprecedented spike in wholesale power 14 costs during the last week of June 15 16 / 17 18 / 19 20 / 21 22 23 24 25 338 Avera, Di-Reb 5A Avista 1 1998 created turmoil in the electricity markets and led 2 certain power marketers to default on their obligations. 3 While this event was temporary, it had a significant 4 negative impact on the 1998 earned returns for individual 5 utilities included in Ms. Carlock's analysis. Value Line 6 noted the transitory impact of these events in its April 7 9, 1999 report on FirstEnergy Corp.: 8 Earnings should be much higher in 1999, with further improvement on tap for 2000. Last year, a 9 surge in purchased-power costs, power marketing losses, and mild weather hurt earnings. We assume 10 no recurrence of these negative factors. (p. 712) 11 Moreover, investors are expecting electric 12 utilities to realize higher rates of return following the 13 transition to competition, as evidenced by the following 14 comments of Merrill Lynch in its Electric Utilities 15 Industry report (January 8, 1997): 16 ... (R)eturns will be limited by the present excess capacity situation, the large amount of 17 current players, and the limited liquidity of the markets. To that end, we believe that these 18 stocks as a group could lose their leadership role for a while. Over the long-term, the players that 19 remain after the capacity situation and number of competitors is rationalized could be very 20 attractive stories. (p. 23) 21 As a result, Ms. Carlock's heavy reliance on recent 22 historical experience caused the results of her 23 comparable earnings analyses to be biased downward. 24 Q. What cost of equity is implied by the data 25 in Ms. Carlock's comparable earnings analyses if a longer 339 Avera, Di-Reb 6 Avista 1 time period is considered? 2 A. As shown on Schedule WEA-7, the rates of 3 return earned on book equity by Moody's electric 4 utilities over the entire 1970-1998 time period shown on 5 Ms. Carlock's Schedule 9 averaged 11.3 percent. This 6 compares with the 8.8 percent 3-year moving average and 7 the 10.7 percent average for 1998 cited by Ms. Carlock 8 (p. 16). Turning to Ms. Carlock's Schedule 10, the rate 9 of return earned on book equity for Moody's gas 10 distribution utilities over this same 29-year period 11 averaged 12.1 percent (Schedule WEA-7). Taken together, 12 13 / 14 15 / 16 17 / 18 19 20 21 22 23 24 25 340 Avera, Di-Reb 6A Avista 1 these more comprehensive analyses of historical earned 2 rates of return demonstrate that Ms. Carlock's data 3 actually imply a "bare bones" cost of equity consistent 4 with the 11.0 to 12.0 percent range I recommended in my 5 direct testimony. 6 Q. Did Ms. Carlock present any other evidence 7 that indicates that the conclusions of her comparable 8 earnings analyses for utilities were biased downward? 9 A. Yes. Ms. Carlock concluded on page 10 of 10 her testimony that: 11 Based upon these considerations my estimate of the near future earned equity capital returns for 12 industrial companies is in the range of 15%-16.0%. 13 With an average long-term Treasury bond yield of 14 approximately 5.8 percent in March 1999, Ms. Carlock's 15 15 to 16 percent cost of equity for industrial firms implies 16 a market equity risk premium ranging from 9.2 to 10.2 17 percent. Viewed in the context of the Capital Asset 18 Pricing Model, multiplying this market equity risk 19 premium by the beta of 0.60 cited by Ms. Carlock (p. 13) 20 produces an equity risk premium for Avista of 5.5 to 6.1 21 percent. Combining this risk premium range with the 5.8 22 percent average yield on long-term treasury bonds results 23 in a cost of equity for Avista based on Ms. Carlock's 24 comparable earnings analyses of 11.3 to 11.9 percent. 25 These calculations are shown below: 341 Avera, Di-Reb 7 Avista 1 Ke= B(Rm-Rf)+ Rf Ke= B(Rm-Rf)+ Rf 2 Ke= 0.60(15%-5.8%)+ 5.8% Ke= 0.60(16%-5.8%)+ 5.8% 3 Ke= 5.5% + 5.8% Ke= 6.1% + 5.8% 4 Ke= 11.3% Ke= 11.9% 5 6 Where: Ke= Cost of Equity 7 B = Beta 8 Rm= Return on Market 9 Rf= Risk-free Rate of Return 10 Q. Please briefly describe Ms. Carlock's DCF 11 analysis. 12 A. Ms. Carlock applied her DCF method to three 13 groups of electric utilities selected from Value Line's 14 Electric Utility-West, Electric Utility-Central, and 15 Electric 16 17 / 18 19 / 20 21 / 22 23 24 25 342 Avera, Di-Reb 7A Avista 1 Utility-East industry groups. She used two dividend 2 yields, one developed by dividing the current annual 3 common dividends declared per share by the average stock 4 price during the past 12 months, and the other based on 5 the dividends actually declared over the past year 6 divided by the average stock price for this same period. 7 In both instances, Ms. Carlock increased her dividend 8 yield by 4.0 percent to allow for equity flotation costs. 9 To the resulting dividend yields, Ms. Carlock added a "g" 10 component based on the average of Value Line's projected 11 growth rates in earnings (EPS) and dividends (DPS) per 12 share over the next 3-5 years. Ms. Carlock went on to 13 conclude that: 14 The cost of equity using the average annual dividend yield for the electric comparable groups 15 produces a range of 10.1%-11.2%. I believe a 10.0% to 11.0% range as the most appropriate 16 estimate under the Discounted Cash Flow method for use in this case. (p. 19) 17 18 Q. How did Ms. Carlock characterize her DCF 19 analysis? 20 A. Ms. Carlock and I are both of the opinion 21 that: 22 ... the constant growth DCF method is not 23 reasonable to use for Avista. (p. 18) But where I 24 elected to apply the non-constant growth DCF model, 25 Ms. Carlock contended that: 343 Avera, Di-Reb 8 Avista 1 The combination of growth estimates with the two-stage DCF method can be just as accurate as 2 the projections of revenue streams and stock prices significantly into the future for use in 3 the non-constant DCF method... I have used the two-stage DCF method with the growth with the two 4 stages averaged for the groups of electric utility comparables. (Pp. 18-19) 5 6 Q. Was Ms. Carlock's application of the DCF 7 model actually a two-stage approach, as she alleged? 8 A. No. As explained on page 48 of my direct 9 testimony, a two-stage DCF model attempts to develop a 10 more realistic growth rate by looking beyond just the 11 near-term and also capturing the longer-term growth 12 expectations investors have likely incorporated into 13 14 / 15 16 / 17 18 / 19 20 21 22 23 24 25 344 Avera, Di-Reb 8A Avista 1 the current prices of common stocks. While I don't 2 disagree with Ms. Carlock's view that the two-stage DCF 3 model "can be just as accurate" as the non-constant 4 growth model I relied on, this will only be the case if 5 it is applied using a growth rate that accurately 6 reflects investors' short- and longer-term expectations. 7 Meanwhile, as Ms. Carlock readily acknowledged, 8 the EPS and DPS growth rates she used were those 9 projected by Value Line for the next 3-5 years. Thus, 10 notwithstanding her contention that she applied a 11 "two-stage" approach, Ms. Carlock's DCF analysis was in 12 fact a constant growth model based solely on near-term 13 projections. As described in Section IV of my direct 14 testimony, investors anticipate that deregulation within 15 the power markets will eventually lead to risks and 16 prospects akin to other firms in the competitive sector 17 and have factored these expectations into the prices of 18 electric utility common stocks, including those used in 19 Ms. Carlock's analysis. Value Line (December 11, 1998) 20 noted in a review of Potomac Electric Power Company, for 21 example, that electric utilities' deregulation and 22 diversification offers the potential to lift long-term 23 growth above near-term expectations: 24 Several nonregulated operations are in the startup phase. PEPCO is investing in the communications, 25 energy efficiency, power marketing, energy 345 Avera, Di-Reb 9 Avista 1 management, gas, and operating and maintenance service businesses. These operations are not 2 likely to contribute much in the next few years, but by 2001-2003 they should begin to lift overall 3 share-net growth closer to management's ultimate 7%-8% goal. (p. 182) 4 5 But in applying the DCF model, Ms. Carlock combined a 6 dividend yield that reflects investors' expectations for 7 a restructured electric utility industry with a 8 short-term growth rate indicative of continued 9 regulation. As a result Ms. Carlock's DCF results 10 understate the rate of return on equity required by 11 investors. 12 Q. How else were Ms. Carlock's DCF results 13 understated? 14 A. Ms. Carlock applied the DCF model: 15 16 / 17 18 / 19 20 / 21 22 23 24 25 346 Avera, Di-Reb 9A Avista 1 ... with projected growth in dividends and projected growth in earnings averaged to use for 2 the growth rate. (p. 19) 3 However, over the near-term period covered by Value 4 Line's projections, electric utilities are widely 5 expected to move their retention ratios towards those of 6 other industries by foregoing dividend increases. S&P 7 noted this divergence between growth in dividends and 8 earnings in a September 26, 1996 Industry Survey: 9 While earnings are on the rise, dividends aren't keeping pace. The reason for this is that more 10 utilities are seeking to increase their cash on hand, so that they can expand or otherwise cope in 11 the face of heightened competition. (p. 5) 12 More recently, Value Line (December 11, 1998) recognized 13 that investors' focus is shifting from dividends to price 14 appreciation as the electric industry becomes more 15 competitive: 16 Going forward, capital gains, as opposed to dividend growth and the yield, will become more 17 important to utility investors. (p. 157) 18 As a result, projected growth in earnings, which 19 ultimately support future dividends and share prices, is 20 likely to provide a more meaningful guide to the growth 21 expectations investors have incorporated into the current 22 prices of electric utility stocks. Nonetheless, 23 near-term projections are apt to understate the long-run 24 growth investors anticipate as regulation is removed and 25 electric utilities' growth begins to approach that of 347 Avera, Di-Reb 10 Avista 1 other competitive firms. 2 Q. What are the results of applying Ms. 3 Carlock's DCF analysis using EPS growth rates? 4 A. As shown on Schedule WEA-8, application of 5 Ms. Carlock's DCF method using projected earnings growth 6 rates produces cost of equity estimates ranging from 7 approximately 10.9 to 12.7 percent. Again, these results 8 support my 11 to 12 percent cost of equity range and 9 indicate that Ms. Carlock's recommended rate of return 10 for Avista is 11 12 / 13 14 / 15 16 / 17 18 19 20 21 22 23 24 25 348 Avera, Di-Reb 10A Avista 1 understated. 2 Q. Do you agree with Ms. Carlock's assessment 3 of Avista's investment risks relative to the other 4 electric utilities she used to estimate the cost of 5 equity? 6 A. No. Ms. Carlock concluded that 7 "(c)ompetitive risks are less for Avista than for most 8 other electric companies" and that Avista's "regulatory 9 risk is low compared to many other regulated utilities" 10 (p. 12), in part based on her observation that: 11 The Idaho Public Utilities Commission has shown overall support for Avista during drought years by 12 providing for surcharges and approving the PCA. (Pp. 12-13) 13 14 While I agree that the PCA moderates the impact of 15 extreme water conditions on Avista's earnings relative to 16 the past, this says nothing about Avista's risk relative 17 to the other electric utilities. As discussed in my 18 direct testimony (Pp. 62-63), the PCA does not eliminate 19 the fluctuations in power production and supply costs 20 associated with Avista's reliance on hydroelectric 21 generation. Thus, the PCA largely addressed a risk 22 factor not faced by most other electric utilities. 23 Because the vast majority of other regulatory 24 jurisdictions already allow the use of some form of 25 adjustment clause to recover power production and supply 349 Avera, Di-Reb 11 Avista 1 costs, the PCA only brings Avista into line with the rest 2 of the electric utility industry. In fact, Value Line 3 has assigned Avista a Regulatory Climate Ranking of 4 "Average", the same as the average for the firms 5 included in Ms. Carlock's Electric Utility B West group. 6 Finally, Schedule WEA-9 presents a comparison of the risk 7 measures Ms. Carlock used to identify her three groups of 8 electric utilities. As shown there, Avista's Beta and 9 Safety Rank are comparable to those maintained by these 10 other firms, while its Timeliness Rank is less favorable. 11 Q. Was Ms. Carlock's discussion of Avista's 12 relative investment risks complete? 13 14 / 15 16 / 17 18 / 19 20 21 22 23 24 25 350 Avera, Di-Reb 11A Avista 1 A. No. Apart from aspects of an electric 2 utility's business risks, such as competitive position 3 and regulatory support, investors are also concerned with 4 the financial risks implied by the amount of debt 5 leverage employed in the capital structure. As explained 6 in Section III of my direct testimony, a lower common 7 equity ratio translates into increased financial risk for 8 all investors. A greater amount of debt and preferred 9 stock means that more investors have a senior claim on 10 available cash flow, which reduces the certainty of 11 receiving contractual payments. From the standpoint of 12 common shareholders, higher debt and preferred stock 13 ratios mean that there are proportionately more investors 14 ahead of them, increasing uncertainty as to the amount of 15 cash flow, if any, that will remain. As a result, common 16 shareholders require a correspondingly higher rate of 17 return to compensate them for bearing the greater 18 financial risk associated with a lower common equity 19 ratio. 20 Q. How does the financial risk implicit in 21 Avista's regulatory capital structure compare with that 22 of the other electric utilities Ms. Carlock referenced in 23 estimating the cost of equity? 24 A. As shown on Schedule WEA-9, the average 25 common equity ratios for Ms. Carlock's three groups of 351 Avera, Di-Reb 12 Avista 1 electric utilities ranged from 45 to 48 percent. 2 Meanwhile, the capital structure Ms. Carlock used to 3 arrive at her overall recommended rate of return for 4 Avista is composed of approximately 37.4 percent common 5 equity. As explained in my direct testimony (Pp. 23-26), 6 Avista's regulatory capital structure falls outside the 7 range maintained by the vast majority of other electric 8 utilities and does not contain sufficient equity to 9 support a single-A bond rating. On the other hand, 10 because there is a greater proportion of lower-cost debt 11 and preferred stock, and correspondingly less higher cost 12 common equity (including associated income taxes), 13 Avista's regulatory capital structure allows ratepayers 14 to benefit 15 16 / 17 18 / 19 20 / 21 22 23 24 25 352 Avera, Di-Reb 12A Avista 1 from a lower effective cost of capital. 2 Q. What does the greater financial risk 3 implicit in Avista's regulatory capital structure imply 4 with respect to Ms. Carlock's recommended rate of return 5 on common equity? 6 A. Investors are risk averse, and will 7 willingly bear additional risk only if they expect 8 compensation. But because Ms. Carlock's cost of equity 9 analysis was predicated on the lower financial risk of 10 her electric utility groups, her recommendation 11 understates the required rate of return associated with 12 Avista's highly leveraged regulatory capital structure. 13 If customers are to benefit from the lower overall cost 14 of capital resulting from Avista's regulatory capital 15 structure, then it is only proper that they bear the 16 higher return on equity required to obtain this benefit. 17 Authorizing a rate of return on equity which ignores the 18 higher risks of Avista's low equity ratio would 19 effectively force shareholders to bear the associated 20 costs while customers would enjoy the benefits. 21 Q. Is there any other evidence that Ms. 22 Carlock's recommended cost of equity is inadequate? 23 A. Yes. Ms. Carlock cited the regulatory and 24 economic standards embodied in the Bluefield and Hope 25 cases (Pp. 3-4) and noted that: 353 Avera, Di-Reb 13 Avista 1 These criteria have been seriously considered in the analysis upon which my recommendations are 2 based. (Pp. 3-4) 3 These judicial precedents require that the fair rate of 4 return on equity should be sufficient to maintain the 5 utility's financial integrity, with bond ratings perhaps 6 providing the most objective guide to a utility's credit 7 standing and ability to attract capital. One of the most 8 important of the quantitative factors determining a 9 utility's bond ratings is its pre-tax interest coverage 10 ratio, which provides a measure of the protection 11 available to pay interest expense 12 13 / 14 15 / 16 17 / 18 19 20 21 22 23 24 25 354 Avera, Di-Reb 13A Avista 1 from operations. 2 The pre-tax interest coverage ratio implied by 3 Ms. Carlock's recommendations is developed on Schedule 4 WEA-10. As discussed in my direct testimony (Pp. 25-26), 5 Avista's capitalization cannot be evaluated directly 6 because of its relatively high reliance on preferred 7 trust securities. Recognizing that these instruments 8 have attributes of both equity and debt securities, S&P 9 has determined that preferred securities can be 10 considered as equity up to 7 percent of Avista's total 11 capital, with the remainder being treated as long-term 12 debt. As shown on Schedule WEA-10, after adjusting 13 Avista's regulatory capital structure to reflect S&P's 14 treatment of preferred securities, Ms. Carlock's 15 recommended rate of return on equity implies a pre-tax 16 interest coverage ratio of 2.63 times. Meanwhile, S&P 17 reported in Utility Financial Statistics (February 1999) 18 that the average pre-tax coverage ratio maintained by 19 other single-A rated electric utilities was 3.61 times. 20 Similarly, the 2.63 times coverage produced by Ms. 21 Carlock's recommended rate of return falls below the 22 benchmark coverage ratio range of 2.75-4.50 times that 23 S&P indicates is necessary to support a single-A bond 24 rating. Thus, Ms. Carlock's rate of return 25 recommendation is insufficient to maintain Avista's 355 Avera, Di-Reb 14 Avista 1 creditworthiness and fails the end result test prescribed 2 by judicial precedent. 3 Q. What then is your conclusion with respect 4 to Ms. Carlock's recommended cost of equity for Avista? 5 A. Ms. Carlock's 10.25 to 11.25 percent cost 6 of equity range understates the rate of return investors 7 require from an investment in the common equity of 8 Avista. At an absolute minimum, Ms. Carlock should have 9 selected a cost of equity from the very top of her range, 10 or 11.25 percent. After incorporating her proposed 25 11 basis point adder for management efficiency, this would 12 imply a fair rate of return on equity for Avista of at 13 least 14 15 / 16 17 / 18 19 / 20 21 22 23 24 25 356 Avera, Di-Reb 14A Avista 1 11.5 percent. In fact, more comprehensive analyses of 2 Ms. Carlock's comparable earnings data and application of 3 her DCF method using more meaningful measures of 4 investors' growth expectations both confirm the 5 reasonableness of the 11.25 to 12.25 percent fair rate of 6 return on equity range for Avista supported in my direct 7 testimony. 8 III. DENNIS E. PESEAU 9 Q. Please summarize Mr. Peseau's 10 recommendations with respect to Avista's rate of return 11 on equity in this case. 12 A. Mr. Peseau recommended a return on equity 13 range for Avista of 10.4 to 10.9 percent. He arrived at 14 this range by adjusting the 12.9 rate of return on equity 15 authorized Avista in 1986 based on changes in the 16 prevailing yields on 30-year treasury bonds. 17 Q. Did Mr. Peseau base his recommendation on 18 any independent analyses of the cost of equity to Avista? 19 A. No. As Mr. Peseau noted in his testimony: 20 ... (M)y analysis is confined to noting the equity return allowed WWP in the 1986 rate case compared 21 to debt costs at that time. (p. 28) 22 Mr. Peseau conducted no independent analyses of a fair 23 rate of return on common equity for Avista. Nor did he 24 make any attempt to ensure that his recommendation is 25 sufficient to compensate investors for the risks to which 357 Avera, Di-Reb 15 Avista 1 they are exposed while maintaining Avista's financial 2 integrity and ability to attract capital. 3 Q. Do changes in long-term Treasury bond 4 yields provide a direct benchmark for adjusting the cost 5 of equity? 6 A. No. While interest rates represent one 7 logical reference point, the impact of fluctuating 8 capital market conditions on the cost of equity is not 9 readily determined. As Mr. 10 11 / 12 13 / 14 15 / 16 17 18 19 20 21 22 23 24 25 358 Avera, Di-Reb 15A Avista 1 Peseau noted: 2 ... I am aware that the spreads between interest rates and equity returns need not be exactly 3 constant over time... (p. 29) 4 In fact, there is substantial evidence that equity risk 5 premiums tend to move inversely with interest rates. In 6 other words, when interest rates rise, equity risk 7 premiums narrow, and when interest rates fall, equity 8 risk premiums are greater. As discussed in Appendix B to 9 my direct testimony, analysis of authorized rates of 10 return on equity for electric utilities indicate that the 11 cost of equity changes approximately one-half as much as 12 the corresponding change in bond yields. Based on this 13 relationship, the 200 to 250 basis-point decline in 14 30-year Treasury bond yields cited by Mr. Peseau would 15 imply a downward adjustment to the cost of equity on the 16 order of 100 to 125 basis points. All else equal, 17 adjusting the 12.9 percent rate of return on equity 18 authorized by the IPUC in 1986 for the decline in 30-year 19 treasury bond yields would imply a current cost of equity 20 on the order of 11.65 to 11.90 percent. Incorporating 21 the 25 basis-point adder for management efficiency 22 recommended by Ms. Carlock results in a current fair rate 23 of return on equity for Avista of 11.90 to 12.15 percent. 24 Q. Apart from the interest rate trends cited 25 by Mr. Peseau, are there other changes that have occurred 359 Avera, Di-Reb 16 Avista 1 that impact investors' required rate of return on equity 2 for electric utilities? 3 A. Yes. Offsetting any impact attributable to 4 changes in interest rates is the ever-increasing 5 uncertainty associated with the restructuring of the 6 electric utility industry. Only since mid-1993 have 7 investors' concerns over the challenges posed by this 8 transition become increasingly magnified. Indeed, 9 Ms. Carlock noted that the risks of utilities and 10 industrial firms were converging (p. 11), concluding 11 that: 12 The competitive risks for gas and electric utilities have changed with the 13 14 15 / 16 17 / 18 19 / 20 21 22 23 24 25 360 Avera, Di-Reb 16A Avista 1 increase in non-utility generation and open transmission access. (Pp. 11-12) 2 3 And as Mr. Peseau granted in his testimony: 4 ... (C)osts or returns on debt and equity tend to move together unless the risk attendant with 5 either changes dramatically. (p. 29, emphasis added) 6 7 Investors' perceptions of the uncertainties associated 8 with electric utilities have undergone a sea change since 9 the IPUC last established the fair rate of return on 10 common equity for Avista. Given the capital markets' 11 current expectations for significantly higher risk, the 12 results of Mr. Peseau's simple-minded, mechanical 13 approach fall well below the rate of return currently 14 required from an equity investment in Avista. 15 Q. Does this conclude your rebuttal testimony 16 in this case? 17 A. Yes, it does. 18 19 20 21 22 23 24 25 361 Avera, Di-Reb 17 Avista 1 (The following proceedings were had in 2 open hearing.) 3 MR. MEYER: The witness is available. 4 COMMISSIONER SMITH: Mr. Ward, do you have 5 questions? 6 MR. WARD: Just a couple. 7 8 CROSS-EXAMINATION 9 10 BY MR. WARD: 11 Q Mr. Avera, if you would turn to your 12 rebuttal testimony at page 16 -- 13 A Yes, sir. 14 Q -- towards the bottom of that page 15 beginning at about line 19 and running on for four or 16 five lines, you talk about the change in risk in the 17 utility industry, which is a subject you touch on from 18 time to time in other portions of your testimony. Do you 19 recall that testimony? 20 A Yes, Mr. Ward. 21 Q And I take it the proposition is that the 22 prospect of restructuring increases risk for utilities 23 and therefore requires -- therefore investors require a 24 somewhat greater return on their investment; is that 25 correct? 362 CSB REPORTING AVERA (X) Wilder, Idaho 83676 Avista 1 A Well, I would say it a little bit 2 differently, that investors are unsure about what the 3 outcome of the restructuring process will be and, 4 therefore, their uncertainty leads them to perceive 5 greater risk and therefore greater return. 6 Q All right. Why would that be true -- well, 7 let me ask it this way: Does it necessarily follow 8 because we are looking at restructuring that all 9 utilities are at risk? 10 A Well, I think, again, our perspective is 11 that of the investors and I think investors believe that 12 all electric utilities will be exposed to a changed 13 operating environment because of the restructuring. I 14 don't think they think every utility will be equally 15 impacted and I think they know there will be differences 16 across the nation, but they are, as I quote numerous 17 times in my testimony, focusing on these changes as a 18 major driver of the risk of investing in this industry. 19 Q If I was considering investing in one of 20 the lowest, if not the lowest, cost investor-owned 21 utilities in the nation, why wouldn't I consider the 22 prospect of a competitive environment for energy sales 23 and specifically for generation and retail sales to be 24 attractive for that utility? 25 A Well, I think a sophisticated investor 363 CSB REPORTING AVERA (X) Wilder, Idaho 83676 Avista 1 would look at the competition the utility faces. We 2 don't have a national electric market. We have a 3 regional electric market, so, for example, in the case of 4 this Company, the local potential competitors are largely 5 public power and co-ops and they have advantageous power 6 supply thanks to the federal government and they have low 7 rates, so I think investors don't look at the absolute 8 national level of rates, they look at the level of rates 9 relative to potential competitors. 10 Q Mr. Avera, have you actually studied the 11 comparative rates of Northwest utilities, publics and 12 privates? 13 A Well, I have looked at the published data 14 for the publics. The data for the privates is not as 15 readily available to people on Wall Street because they 16 can't follow them, don't look at them. From talking to 17 Company officials, some of the localities, some of the 18 co-ops like, I think it's, Northern Lights, some of the 19 public utility districts have very favorable costs 20 relative to Avista and depending on how federal 21 legislation works out may be even more favorably 22 impacted. 23 Q One last question. If in fact we are to 24 increase returns and, therefore, rates on the grounds 25 that the prospect of restructuring requires compensation 364 CSB REPORTING AVERA (X) Wilder, Idaho 83676 Avista 1 for this presumed risk, but in a state -- wouldn't it be 2 true in a state that does not intend restructuring the 3 ratepayers get the worst of both worlds; that is, they 4 get to pay a somewhat higher rate because of the prospect 5 of restructuring, but, of course, they don't get access 6 to the benefits, if any, of restructuring? 7 A Well, I think -- 8 Q Isn't that true? 9 A I would answer the question as you started 10 to answer it, Mr. Ward. I don't think it's necessarily 11 the worst of both worlds because I think at this point 12 the benefits of restructuring are problematic and I think 13 that's the reason that restructuring is proceeding at a 14 different pace in different states, because some states 15 see greater benefit than others, so I think the benefits 16 are a function of how the state tailors its regulatory 17 environment to the needs and circumstances of the state, 18 but the state cannot control the requirements of 19 investors. 20 Investors, if they're going to put money 21 into this business, have to be convinced they're 22 adequately compensated and in the current environment, 23 given all of the uncertainties about state, federal 24 legislation, technology, investors require rate of 25 returns. 365 CSB REPORTING AVERA (X) Wilder, Idaho 83676 Avista 1 Q But in any event, under the theory that 2 you've proposed, the long and short of it is for 3 ratepayers rates go up? 4 A Well, the cost of money is just one of the 5 elements in the cost of service, but I think the reality 6 is, and I think this is what Mr. Dr. Peseau failed to 7 realize, is there has been a sea change in the relative 8 risk of the utility industry in the last several years. 9 MR. WARD: I'll leave it at that. Thank 10 you. 11 COMMISSIONER SMITH: Mr. Shurtliff. 12 MR. SHURTLIFF: Yes, thank you. 13 14 CROSS-EXAMINATION 15 16 BY MR. SHURTLIFF: 17 Q At page 64 of your direct testimony, you 18 indicate that you after doing the magic that you do to 19 get there come to a conclusion that a fair rate of return 20 on common equity for Washington Water Power or WWP would 21 be between 11.25 to 12.25 percent. Is that your 22 conclusion? 23 A After I do my professional analytical work, 24 the result is the Avista return should be in that range. 25 Q All right. How did you describe that, 366 CSB REPORTING AVERA (X) Wilder, Idaho 83676 Avista 1 professional analytical work? 2 A Yes, sir. 3 Q And then what did you apply to get from 4 11.25 to 11.75, more professional analytical work or was 5 that just a judgment that you ought to come someplace in 6 the midpoint range? 7 A It was a judgment that I lay out in my 8 testimony to go to the midpoint because that seemed to be 9 a logical way of handling this range. I think just like 10 Ms. Carlock, you can't resolve these down to particular 11 numbers, you come up with a range and then you suggest a 12 range within which the Commission might consider their 13 decision. 14 Q And so if the Commission concluded 11.25 15 equity was appropriate, they wouldn't be, as you 16 characterized Dr. Peseau, simpleminded in that regard? 17 A Absolutely not. I think they would be 18 within the range and I think that would be appropriate. 19 Let me quickly say that I was speaking of other things 20 about Dr. Peseau, not his use of the range. 21 Q You were speaking of his professional 22 analytical work, were you not, as simpleminded? 23 A No, sir. I was speaking of the absence 24 thereof, the fact that he only looked at the change in 25 interest rates between the last Order and did not 367 CSB REPORTING AVERA (X) Wilder, Idaho 83676 Avista 1 consider the fundamental changes in the industry, nor did 2 he do any independent analytical work trying to derive 3 the investors' required return. 4 Q Also, on page 64 you conclude that the 5 midpoint range that you found, 11.75, should be bolstered 6 by the .25 basis points for good work, for a job well 7 done. 8 A Well, I say that that, I believe, is 9 appropriate regulatory policy. I am not -- Mr. Dukich is 10 the one that describes the nature of the performance of 11 the Company that merits the 25 basis points. My role in 12 this is to say that that is an appropriate way to treat 13 my 11.75 is to add 25 basis points to it. 14 Q If Mr. Dukich's predicates are correct. 15 You didn't do your own analysis in that regard as to 16 whether 25 basis points were appropriate? 17 A No, sir, I did not. 18 Q Would it be fair to say that you analyzed 19 what Dr. Peseau did and you analyzed what Ms. Carlock did 20 and you've just indicated, as I quoted from page 2 of 21 your rebuttal, you indicate a judgment that Dr. Peseau's 22 was a simpleminded approach, you just disagree with 23 Ms. Carlock, I take it? 24 A Well, I disagree with a lot of inferences 25 she draws from her numbers. I think in fact the 368 CSB REPORTING AVERA (X) Wilder, Idaho 83676 Avista 1 empirical evidence she presents corroborates my range and 2 I agree with many of the observations she makes about the 3 changes in the industry, about the changing relative risk 4 and about the inadequacy of the single growth rate, 5 discounted cash flow model from capturing required 6 returns in this dynamic industry at this point. 7 Q So as I understand it, you didn't have any 8 difficulty with her approach, it was her conclusion that 9 you find fault with? 10 A That is correct, her bottom line numbers. 11 I didn't disagree with her methodological approaches. I 12 just disagreed with some of the steps she took along the 13 way and I think if she had taken different steps, she 14 would have come up with essentially the same result as I 15 did. 16 Q Did she apply professional analytical work 17 to get to where she got? 18 A Yes, sir, she did. 19 Q Would it be fair to say that among experts 20 in your field or any other that different experts can 21 look at the same set of facts or predicates and come to 22 different conclusions having exercised due and reasonable 23 professional analytical work? 24 A Yes, sir, that's why it's important for us 25 to lay out our numbers and our reasoning in our testimony 369 CSB REPORTING AVERA (X) Wilder, Idaho 83676 Avista 1 so the decision makers can make a weighing of our 2 analytical judgments. 3 Q And that's basically what the Commission is 4 required to do, is it not, in the instance of cost of 5 equity or any number of the other parameters in rate 6 cases is they have to sort out what the experts tell them 7 and apply their professional analytical work to it to see 8 if the basis or bases of the experts' work justifies the 9 conclusion that they come to? 10 A Yes, sir. 11 Q And so they could come to a different 12 conclusion than you enunciate and still be correct, could 13 they not? 14 A They do sadly, sometimes they do. 15 MR. SHURTLIFF: I feel that way about 16 judges quite often. Thank you. 17 THE WITNESS: Thank you, sir. 18 COMMISSIONER SMITH: Mr. Woodbury. 19 MR. WOODBURY: Thank you, Madam Chair. 20 21 22 23 24 25 370 CSB REPORTING AVERA (X) Wilder, Idaho 83676 Avista 1 CROSS-EXAMINATION 2 3 BY MR. WOODBURY: 4 Q Mr. Avera, how are you? 5 A I'm fine, and you, sir? 6 Q Good. When investors review financial 7 statements and rating agency reports, isn't it true that 8 the capital structure reviewed is the reported ratios 9 based on the books? 10 A That's the starting point. Very often 11 rating agencies make adjustments, like in my rebuttal 12 testimony I point out that Standard and Poor's makes an 13 adjustment for Avista's preferred stock to move some of 14 it down to equity, so they start with the reported 15 financial numbers, but then they make adjustments that 16 they think are more indicative of the risk. 17 Q Would you agree that hydro generation is 18 considered low-cost energy? 19 A Generally, yes, sir. 20 Q And would you also agree that Avista is 21 available to invest its retained earnings in nonregulated 22 operations? 23 A Yes, sir. That's something that Avista and 24 most other electric utilities in the country are doing 25 now. 371 CSB REPORTING AVERA (X) Wilder, Idaho 83676 Avista 1 MR. WOODBURY: Thank you. 2 Madam Chairman, no further questions. 3 COMMISSIONER SMITH: Do we have questions 4 from the Commissioners? 5 Redirect, Mr. Meyer? 6 MR. MEYER: Just one or two brief 7 redirect. 8 9 REDIRECT EXAMINATION 10 11 BY MR. MEYER: 12 Q You indicated in response, I believe, to 13 intervenor's examination that in fact some of the work 14 done by Ms. Carlock on behalf of Staff actually 15 corroborates your proposed range of reasonableness for 16 return on equity. Can you give me an example of a type 17 of adjustment that if made, let's say, to capital 18 structure would bring the two proposed ranges into sync? 19 A Yes. One of my disagreements with 20 Ms. Carlock's work is she fails to consider that for 21 regulatory purposes, Avista has requested a much smaller 22 equity ratio than the rating agencies look at and that 23 adjustment alone is worth about 75 basis points in a cost 24 of equity; in other words, had Avista used a capital 25 structure like Idaho Power, for example, Avista would 372 CSB REPORTING AVERA (Di) Wilder, Idaho 83676 Avista 1 have gotten an -- would need an equity return of 75 basis 2 points to offset the difference in capital structure. 3 Q And in preparing your testimony today, did 4 you review prior testimony of Ms. Carlock in the Idaho 5 Power case? 6 A Yes. Because I had been involved in that 7 case, I had a copy of her testimony in my archives, so I 8 pulled that out and noted what the capital structure had 9 been in that case. 10 Q Very briefly, compare the two cap 11 structures and the resulting return. 12 MR. WOODBURY: Madam Chair? 13 COMMISSIONER SMITH: Mr. Woodbury. 14 MR. WOODBURY: Perhaps Mr. Meyer could 15 indicate how this is redirect. I don't recall a line of 16 questioning along these -- 17 COMMISSIONER SMITH: Mr. Meyer. 18 MR. MEYER: We're merely trying to 19 illustrate the point about how the two expert witnesses 20 in this area are not that far apart if you just make the 21 necessary adjustment, in effect corroborating his 22 statement that Staff's position really is in sync to the 23 Company's case. I'm giving an example of that. 24 COMMISSIONER SMITH: Is this something in 25 his rebuttal? 373 CSB REPORTING AVERA (Di) Wilder, Idaho 83676 Avista 1 MR. MEYER: No, but it's in response to 2 examination by Mr. Shurtliff. That's all. I don't mean 3 to prolong it, but I'm just trying to cite an example. 4 COMMISSIONER SMITH: Please proceed. 5 THE WITNESS: I looked at Ms. Carlock's 6 Schedule 13 which is essentially the same as her 7 Schedule 14 in the testimony in this case where she 8 presents the capital structure, the component cost and 9 then comes down to a weighted average cost of capital, 10 and basically if you took the cost in the Idaho Power 11 case but put in the capital structure that Avista is 12 asking for in this case, to get to the same weighted 13 average cost of capital, instead of an 11 percent return 14 on equity, you would have an 11.71 percent return on 15 equity or you can go the other way and take Avista's 16 capital structure in this case and work backward with the 17 component cost and see what the cost of equity or what 18 the weighted average cost would be if we had used Idaho 19 Power's capital structure and the answer is if you use 20 11.25 percent, which is the top of Ms. Carlock's cost of 21 equity range, you get essentially the same weighted 22 average cost of capital that I recommended using a 12 23 percent return, so 11.25 when you apply it to recognize 24 the difference in capital structure is equivalent to a 12 25 percent return, so my range and her range are really the 374 CSB REPORTING AVERA (Di) Wilder, Idaho 83676 Avista 1 same once you make the capital structure adjustment. 2 MR. MEYER: Thank you. That's all I have. 3 COMMISSIONER SMITH: Okay, thank you. 4 Thank you, Mr. Avera. 5 THE WITNESS: Thank you. 6 (The witness left the stand.) 7 MR. MEYER: Let's see, the next witness in 8 order is Mr. Norwood. 9 10 KELLY O. NORWOOD, 11 produced as a witness at the instance of Avista 12 Corporation, having been first duly sworn, was examined 13 and testified as follows: 14 15 DIRECT EXAMINATION 16 17 BY MR. MEYER: 18 Q Are you ready? 19 A I'm ready. 20 Q Please for the record state your name and 21 your employer. 22 A My name is Kelly Norwood. I work for 23 Avista Corp. 24 Q And have you prepared both direct and 25 rebuttal testimony? 375 CSB REPORTING NORWOOD (Di) Wilder, Idaho 83676 Avista 1 A Yes, I have. 2 Q Do you have any material changes to make to 3 that? 4 A I do have one correction on page 16 of my 5 direct testimony. 6 Q Okay, give us a minute to turn to that. 7 Page 16? 8 A Page 16 in my direct testimony beginning on 9 line 9. 10 Q Go ahead. 11 A There are three numbers listed there, 12 weighted average market prices. These numbers that are 13 listed there inadvertently were simple average numbers. 14 I need to change those to the weighted average numbers. 15 The 20.70 should be changed to 21.78; the 14.82 should be 16 changed to 14.53; and the 12.79 should be changed to 17 13.25, and again, those are the weighted average numbers 18 to compare to pro forma numbers. 19 Q Okay, any other changes to your direct or 20 rebuttal? 21 A None. 22 Q So if I were to ask you the questions that 23 appear therein, with those changes having been made, 24 would your answers be the same? 25 A Yes. 376 CSB REPORTING NORWOOD (Di) Wilder, Idaho 83676 Avista 1 Q You're sponsoring Exhibits 6 through 9 and 2 Exhibit 23, aren't you? 3 A That's correct. 4 Q Any changes to those? 5 A None. 6 Q And is the information contained therein 7 true and correct? 8 A Yes. 9 MR. MEYER: With that, I ask that 10 Mr. Norwood's direct and rebuttal be spread into the 11 record as if read and move for the admission of Exhibits 12 6 through 9 as well as 23. 13 COMMISSIONER SMITH: If there's no 14 objection, the testimony will be spread upon the record 15 as if read and Exhibits 6 through 9 and 23 will be 16 admitted. 17 MR. MEYER: Thank you. 18 (Avista Corporation Exhibit Nos. 6 - 9 19 and 23 were admitted into evidence.) 20 (The following prefiled direct and 21 rebuttal testimony of Mr. Kelly Norwood is spread upon 22 the record.) 23 24 25 377 CSB REPORTING NORWOOD (Di) Wilder, Idaho 83676 Avista 1 I. INTRODUCTION/SUMMARY 2 Q. Please state your name, the name of your 3 employer and your business address. 4 A. My name is Kelly O. Norwood. I am employed 5 by The Washington Water Power Company (to be renamed 6 Avista Corporation, effective January 1, 1999) at 1411 7 East Mission Avenue, Spokane, Washington. 8 Q. In what capacity are you employed? 9 A. I am employed as a Regulatory Policy 10 Analyst in the Rates Department. 11 Q. Please state your educational background 12 and professional experience. 13 A. I graduated from Eastern Washington 14 University in 1981 with a Bachelor of Arts Degree in 15 Business Administration, majoring in Accounting. I was 16 hired by the Company in June 1981. Over the past 17 17 years I have spent eight years in the Rates Department 18 with involvement in cost of service, rate design, revenue 19 requirements, demand side management, electric industry 20 restructuring, and the design and implementation of 21 retail access pilot programs. I have also spent nine 22 years in the power supply department (currently Resource 23 Optimization) negotiating power contracts, preparing 24 economic analyses of resource alternatives, modeling the 25 costs associated with the Company's hydroelectric and 378 Norwood, Di 1 WWP 1 thermal resource operations, and managing the risks 2 associated with energy trading and marketing operations. 3 I moved from the Resource Optimization Department to the 4 Rates Department in November 1998. 5 Q. What is the scope of your testimony in this 6 proceeding? 7 A. My testimony will explain the adjustments 8 to the 1997 test period power supply revenues and 9 expenses. My testimony will also discuss the Company's 10 Dispatch Simulation Model (Dispatch Model) and how it is 11 used by the Company to normalize power costs that are 12 dependent upon streamflow and wholesale market conditions 13 for ratemaking purposes. 14 15 / 16 17 / 18 19 / 20 21 22 23 24 25 379 Norwood, Di 1A WWP 1 Q. Are you sponsoring exhibits in this 2 proceeding? 3 A. Yes. I am sponsoring four exhibits that 4 are marked as Exhibit Nos. 6 through 9 for 5 identification. 6 Q. Please identify the specific power supply 7 revenue and expense items that are covered by your 8 testimony and the total adjustment being proposed. 9 A. Column (a) on Pages 1 through 4 of Exhibit 10 No. 6 identifies the power supply revenue and expense 11 items that fall within the scope of my testimony. These 12 revenue and expense items are, in general, related to 13 long-term and short-term wholesale purchases and sales, 14 wheeling transactions, thermal fuel expenses, as well as 15 other miscellaneous power supply related expenses. Any 16 adjustments to other power supply related revenues and 17 expenses will be addressed by other witnesses. 18 Column (b) shows the 1997 test period actual 19 revenues and expenses for each of the items listed. 20 Column (c) includes the adjustments to the 1997 test 21 period revenues and expenses. Column (d) shows the pro 22 forma level of power supply revenues and expenses that 23 are being proposed by the Company in this case. 24 Line 130, Column (c), on Page 3 shows a net increase 25 in 1997 actual power supply revenues and expenses of 380 Norwood, Di 2 WWP 1 $46,764,000 on a system basis. The Idaho jurisdictional 2 share of the system number, using a 33.18% allocation 3 percentage, is $15,516,000. 4 Q. What is the basis for making the 5 adjustments to the 1997 actual revenues and expenses? 6 A. Adjustments are made to normalize any 7 revenue and expense items that were abnormal during the 8 test period. For example, accounting entries during the 9 1997 test period that included dollar amounts that relate 10 to a prior period are eliminated. To the extent that 11 revenue 12 13 / 14 15 / 16 17 / 18 19 20 21 22 23 24 25 381 Norwood, Di 2A WWP 1 and expense items are influenced by weather, streamflow 2 and wholesale market conditions, adjustments are made to 3 set the revenues and expenses based on normal weather, 4 and normal streamflow and wholesale market conditions. 5 In addition, adjustments are made to reflect known 6 and measurable changes between the 1997 test period and 7 the time period that retail rates would be in effect (the 8 pro forma period). For example, power contracts that 9 terminate at the end of the 1997 test period will not be 10 in place when a Commission order is issued in this case 11 and new rates are implemented. Therefore, adjustments 12 are made for known and measurable power supply revenue 13 and expense changes so that the proper costs are 14 reflected in customers' rates at the time they are 15 implemented. 16 In this filing the Company has included pro forma 17 power supply adjustments to reflect power costs for the 18 twelve-month period beginning July 1, 1999 and ending 19 June 30, 2000. 20 Q. Have these power supply adjustments been 21 prepared using the same methods that have been used in 22 prior general rate cases? 23 A. Yes. 24 Q. What are the major power supply changes 25 that contribute to the total power supply adjustment of 382 Norwood, Di 3 WWP 1 $46,764,000 on a system basis? 2 A. The power supply revenue and expense items 3 identified below represent the major changes that 4 contribute to the total power supply adjustment of $46.8 5 million. Although there are many other revenue and 6 expense items that include adjustments, these items 7 provide an indication of the major components of the 8 total power supply adjustment. 9 Hydroelectric Generation: Streamflow during the 1997 test period was among the highest 10 on record. The January through July runoff at The Dalles on the Columbia River was 150% of normal, 11 and flows on the Clark Fork River were 169% of normal. During 1997 the Company implemented two 12 Power Cost Adjustment rebates to its Idaho retail customers related to favorable streamflow 13 conditions. Hydroelectric generation from the 14 15 / 16 17 / 18 19 / 20 21 22 23 24 25 383 Norwood, Di 3A WWP 1 Company's system, and its share of the Mid-Columbia generation, for 1997 was 141 average 2 megawatts above normal. An estimate of the impact of a normalization adjustment related solely to 3 hydroelectric generation is an increase in net expense of $18.3 million. 4 NCPA Contract: The net increase in expense from 5 the termination of the power sale contract to NCPA effective December 1997 is approximately $6.2 6 million. 7 PP&L 1989 Contract: The net increase in expense from the termination of the power sale contract to 8 PacifiCorp effective December 1997 is approximately $5.9 million. 9 Puget Contract: The decrease in the rate for the 10 power sales contract to Puget Sound Energy from $32.11/Mwh to $23.20/Mwh effective April 1, 1997 11 results in a net increase in expense of approximately $1.5 million. 12 Potlatch Contract: The increase in the rate for 13 the contract to purchase power from Potlatch Corporation from $43.96/Mwh to $46.70/Mwh results 14 in a net increase in expense of approximately $1.3 million. 15 Wheeling Contracts: Wheeling revenues associated 16 with the BPA Borderline Loads and the Cogentrix contract decrease by $2.5 million. 17 Enron Exchange: Termination of the exchange with 18 Enron effective September 1998 results in a net increase in expense of approximately $1 million. 19 Kettle Falls Fuel: An increase in the cost per 20 ton for Kettle Falls fuel results in a net increase in expense of approximately $1 million. 21 Retail Loads: Retail loads for the pro forma 22 period, compared to the test period, are 24 aMw higher due to the termination of the DADS pilot 23 program (18 aMw), and the weather normalization adjustment (6 aMw). These load adjustments 24 increase net power supply costs by approximately $3.1 million. (An increase in retail revenues has 25 been reflected in the exhibits of other company witnesses associated with these retail load 384 Norwood, Di 4 WWP 1 adjustments.) 2 Wholesale Market Price: The pro forma weighted average wholesale market price is $18.32/Mwh 3 compared with a relatively low 1997 test period actual of $14.82/Mwh, due primarily to near-record 4 streamflow conditions. As a net purchaser of short-term power during the pro forma period of 5 approximately 50 aMw, the net increase in expense related solely to the market price is 6 approximately $1.5 million. 7 8 / 9 10 / 11 12 / 13 14 15 16 17 18 19 20 21 22 23 24 25 385 Norwood, Di 4A WWP 1 Detailed work papers have been provided to the 2 Commission Staff, simultaneously with this filing, that 3 support each of the pro forma adjustments on Exhibit 4 No. 6, along with a brief description of each adjustment. 5 6 II. NORMALIZATION OF POWER COSTS RELATED TO STREAMFLOW AND WHOLESALE MARKET CONDITIONS 7 8 Q. Why are the test period revenues and 9 expenses adjusted to reflect normal streamflow and 10 wholesale market conditions? 11 A. The actual short-term sales, short-term 12 purchases, and thermal fuel expenses during the 1997 test 13 period were dependent on the market conditions that 14 existed during that period, and the hydroelectric 15 generation from actual streamflow. The actual streamflow 16 for the 1997 test period was well above normal and could 17 not be expected to reoccur each year. The Company's base 18 retail rates are established using normal or average 19 streamflow conditions so that, over time, the base retail 20 rates would fully recover the Company's power supply 21 costs. It is necessary, therefore, to adjust the 1997 22 actual level of revenues and expenses to reflect normal 23 streamflow conditions and current wholesale market 24 conditions. 25 Q. How does the Company's Power Cost 386 Norwood, Di 5 WWP 1 Adjustment (PCA) mechanism relate to the power supply 2 costs that are included in base retail rates under normal 3 streamflow and wholesale market conditions? 4 A. The PCA mechanism is designed to track the 5 changes in power costs, from the normalized level, that 6 are directly related to streamflow and wholesale market 7 conditions. To the extent that the Company's actual 8 hydroelectric generation and the actual wholesale market 9 prices are different than the normalized levels 10 established in this case, future changes in power supply 11 12 / 13 14 / 15 16 / 17 18 19 20 21 22 23 24 25 387 Norwood, Di 5A WWP 1 costs associated with these variations will be quantified 2 and deferred for later rebate or surcharge to the 3 Company's retail customers. This true-up is done on a 4 monthly basis with deferrals made to a PCA balancing 5 account. When the balancing account reaches $2.2 million 6 in either the rebate or surcharge direction, a temporary 7 retail rate adjustment is made through Tariff Schedule 8 66-Temporary Power Cost Adjustment-Idaho to implement the 9 rebate or surcharge. The Company's base retail rates, 10 however, remain unchanged through this process. 11 The power supply costs approved by the Commission in 12 this case will establish the new level of normalized 13 power costs. These normalized power costs will be used 14 by the Company in future PCA calculations to measure 15 changes in power costs related to variations in 16 hydroelectric generation and wholesale market conditions. 17 18 III. DISPATCH SIMULATION MODEL OPERATION AND INPUTS 19 Q. Please explain how the power supply 20 revenues and expenses that are dependent upon streamflow 21 and wholesale market conditions are adjusted to reflect 22 normal conditions? 23 A. The Company's Dispatch Simulation Model 24 (Dispatch Model) is used to normalize power supply 25 revenues and expenses that are dependent upon streamflow 388 Norwood, Di 6 WWP 1 and wholesale market conditions, including short-term 2 sales, short-term purchases and thermal fuel expenses. 3 The model was designed to simulate, as closely as 4 possible, the operation of the Company's available 5 resources to serve retail load and wholesale obligations 6 under varying hydroelectric and wholesale market 7 conditions. 8 Q. In general terms, how does the Dispatch 9 Model work and what are the inputs? 10 A. Exhibit No. 7 includes a schematic that 11 illustrates the operation of the Dispatch Model. The 12 Dispatch Model develops power supply costs under varying 13 hydroelectric conditions 14 15 / 16 17 / 18 19 / 20 21 22 23 24 25 389 Norwood, Di 6A WWP 1 by modeling the relationship of energy supply, energy 2 demand and wholesale market prices in the Northwest. 3 This is done by assigning different prices to different 4 levels of surplus energy available in the Northwest under 5 60 different annual streamflow conditions. These 6 regional surplus conditions and the corresponding 7 regional market prices are input into the Dispatch Model 8 along with the generation from the Company's own 9 hydroelectric projects under these 60 different 10 streamflow conditions (720 months). In addition, the 11 available generation from the Company's thermal plants, 12 and the Company's retail load and firm contract rights 13 and contract obligations are input. 14 As illustrated on Exhibit No. 7, the Dispatch 15 Model then simulates the operation of the Company's 16 thermal plants, and the short-term sales and purchases 17 that would occur, under each of the 60 different annual 18 hydroelectric and market conditions, to serve retail load 19 and contract obligations. The average results from the 20 60 streamflow and market conditions are used as the 21 normalized level of short-term sales, short-term 22 purchases, and thermal fuel expenses. 23 Q. What source is used for the Company's 24 hydroelectric generation and regional surplus for the 60 25 historical streamflow conditions? 390 Norwood, Di 7 WWP 1 A. The regional surplus energy values were 2 obtained from studies prepared by the Northwest Power 3 Pool Coordinating Group (NWPP) for the 1996-97 operating 4 year. The NWPP prepares a Headwater Benefits Study each 5 year to determine the benefits to down-stream parties 6 from up-stream storage releases. They also prepare what 7 is called a "critical period" study to determine 8 operating rule curves for the operation of Northwest 9 hydroelectric storage reservoirs for the upcoming 10 operating year. 11 The NWPP Coordinating Group receives data submittals 12 each year from members of the NWPP, which includes most 13 of the investor-owned and public utilities in the 14 Northwest. These 15 16 / 17 18 / 19 20 / 21 22 23 24 25 391 Norwood, Di 7A WWP 1 data submittals include the retail loads, wholesale 2 contract rights and obligations, and available thermal 3 and hydroelectric generating capability for each of the 4 parties for the upcoming operating year. The NWPP runs a 5 hydro regulation model that simulates the operation of 6 Northwest reservoirs, including Canadian reservoirs, 7 based on the loads to be served and the available 8 hydroelectric and thermal resources for the Northwest 9 region. This simulation of reservoir operation is based 10 on streamflow that occurred in each of the 60 historical 11 operating years from July 1, 1928 to June 30, 1988. The 12 results from the study include the amount of energy that 13 could be produced at all Northwest hydroelectric projects 14 for each month under each of these streamflow conditions, 15 and the amount of surplus energy that would be available 16 in the Northwest. 17 Thus, the NWPP study shows what the output would be 18 from the Company's existing hydroelectric projects, under 19 the current rules for reservoir operations, for 60 20 different annual streamflow conditions. The study also 21 shows the amount of surplus that would be available in 22 the Northwest for each month under these 60 different 23 streamflow conditions. The 60 years of monthly 24 hydroelectric generation for the Company's hydro projects 25 are input into the Dispatch Model, along with the 392 Norwood, Di 8 WWP 1 corresponding regional (Northwest) surplus numbers that 2 are used to model wholesale market prices. 3 Q. Why does the Company use 60 years of 4 historical streamflow conditions for normalization 5 instead of some other number? 6 A. Normalization of short-term sales, 7 short-term purchases and thermal fuel expenses is based 8 on an average of streamflow and market price conditions 9 over time. Absent a trend or pattern in the data, use of 10 the maximum amount of available, reliable historical 11 streamflow data will yield the most accurate results. 12 The Company is not aware of any proven trend or pattern 13 in 14 15 / 16 17 / 18 19 / 20 21 22 23 24 25 393 Norwood, Di 8A WWP 1 the historical data. The 60 years of streamflow data in 2 the NWPP study is generally accepted in the region as 3 reliable data, and is used by many in the region for 4 studies related to hydroelectric operations. 5 This streamflow data is updated once every ten 6 years, at which time an additional ten years of 7 streamflow data is added. When the data is updated each 8 ten years, the historical streamflow for each tributary, 9 for each year beginning July 1, 1928, are adjusted to 10 reflect the most recent levels of irrigation and 11 depletion levels. This requires a significant time 12 commitment, which is why the data is updated only once 13 every ten years. In the most recent update, completed in 14 1993, the Bonneville Power Administration (BPA) hired 15 A.G. Crook Company, as a consultant, to coordinate the 16 effort. The consultant enlisted the assistance of 17 Northwest utilities and various state and federal 18 agencies in updating the data, including the U.S. Army 19 Corp of Engineers, and the U.S. Bureau of Reclamation. 20 Q. How are the regional surplus numbers from 21 the NWPP study used to model wholesale market prices in 22 the Dispatch Model? 23 A. As stated earlier, the Dispatch Model 24 assigns a different market price to the different levels 25 of surplus energy available in the Northwest. This is 394 Norwood, Di 9 WWP 1 accomplished through the use of six "bands" in the 2 Dispatch Model for regional surplus. Each band is 3 assigned a separate market price and has a specific size 4 in average megawatts. The bands represent the markets 5 available for Northwest surplus energy and are ranked 6 from highest priority to lowest priority as shown on 7 Exhibit No. 8. The price assigned to a band is used as 8 the wholesale market price in the Dispatch Model when the 9 regional surplus value falls within that band. For 10 example, a regional surplus of 500 aMw for the month of 11 July would fall in Band 1, and the market price under 12 that regional surplus condition would be $35.00/Mwh, as 13 indicated on Exhibit No. 8. A 14 15 / 16 17 / 18 19 / 20 21 22 23 24 25 395 Norwood, Di 9A WWP 1 regional surplus of 1,400 aMw in July would exceed the 2 Band 1 limit, and is less than the Band 2 limit, and 3 therefore would fall in the second band with a market 4 price of $23.00/Mwh. The market price assigned to the 5 first band corresponds to a small regional surplus. As 6 the regional surplus increases, the corresponding market 7 price decreases. 8 Q. Please describe each of the six bands and 9 how they were developed. 10 A. The majority of the data used to develop 11 the six bands was taken from the Headwater Benefits Study 12 prepared by the NWPP for the 1996-97 operating year. 13 This study determines the benefits to down-stream parties 14 from up-stream storage releases, based on the operation 15 of the reservoirs in the NWPP hydro regulation model for 16 the relevant operating year. 17 In the Headwater Benefits Study, the NWPP identifies 18 potential markets for the regional surplus that comes 19 from the NWPP study. These markets can include 20 interruptible energy sales to parties within the 21 Northwest (interruptible loads are generally not included 22 in the NWPP study as firm loads and, thus, must be served 23 with surplus power), sales of energy to parties outside 24 the region, and displacement of thermal generation. 25 The first of the six market bands is labeled 396 Norwood, Di 10 WWP 1 "Primary Requirements." This band includes DSI 2 interruptible loads as shown on Line 1 of Exhibit No. 8 3 and displaceable high-cost oil-fired and gas-fired 4 thermal generation on Lines 2-7. When there is a 5 relatively small amount of surplus in the region, as in a 6 Band 1 condition, the market price will likely be 7 determined by the incremental cost of oil and gas-fired 8 generation, as well as purchases of energy from the 9 Southwest. Efforts would be made to serve the DSI 10 interruptible loads. The Band 1 price is set at 11 $35.00/Mwh based on the estimated incremental costs for 12 these high-cost generating units, and the estimated 13 purchase price for energy from the Southwest. In the 14 Dispatch Model, when the 15 16 / 17 18 / 19 20 / 21 22 23 24 25 397 Norwood, Di 10A WWP 1 regional surplus falls within the Band 1 limit, e.g., at 2 or below 1,241 aMw for July, the market price is set at 3 $35.00/Mwh. 4 Band 2 is labeled "High Cost Thermal." The amount 5 of displaceable Beaver and Rathdrum generation is 6 included in this band. Both of these projects are 7 gas-fired plants, but the heat rates are lower than the 8 plants in Band 1 and, thus, have a lower incremental 9 cost. The price assigned to Band 2 is $23.00/Mwh based 10 on the approximate incremental costs of these units. 11 When the regional surplus exceeds the Band 1 limit, but 12 is less than the Band 2 limit, the market price in the 13 Dispatch Model for the respective month is set at 14 $23.00/Mwh. 15 Band 3 is labeled "Primary Markets." The Southwest 16 Market is included on Line 14 and reflects the 17 opportunity for sales of Northwest surplus energy to the 18 Southwest on the Pacific Northwest-Southwest Interties. 19 The intertie capabilities have been adjusted to reflect 20 maintenance outages, loop flow and firm contract 21 obligations. The other market opportunities for 22 Northwest surplus in Band 3 include the displaceable 23 energy from a number of thermal plants, as shown on 24 Exhibit No. 8, from the NWPP Headwater Benefits Study. 25 These plants have been grouped together in Band 3 because 398 Norwood, Di 11 WWP 1 they have similar incremental operating costs. The Band 2 3 market price has been set at $20.00/Mwh based on the 3 incremental operating costs of these plants, and the 4 estimated price at which the Southwest would make major 5 purchases of Northwest surplus. 6 Band 4 is "High Cost Displacement." As with the 7 other bands, the thermal plants in Band 4 have been 8 grouped together because they have similar incremental 9 operating costs. The displaceable generation for 10 Centralia has been split between Band 4 and Band 5 to 11 reflect the difference in incremental fuel costs 12 associated with external coal imported from the Powder 13 14 / 15 16 / 17 18 / 19 20 21 22 23 24 25 399 Norwood, Di 11A WWP 1 River Basin (Band 4), and the native coal from the local 2 mine (Band 5). The market price for Band 4 has been set 3 at $13.00/Mwh. 4 Band 5 is labeled "Low Cost Displacement" and 5 includes the Eastern Market and the displaceable 6 Centralia generation fueled by native coal. The Eastern 7 Market reflects the opportunities to sell Northwest 8 surplus to the east through Idaho and Utah to displace 9 low cost thermal generation, e.g., the PacifiCorp/Utah 10 thermal plants. The Band 5 price has been set at 11 $10.00/Mwh. 12 Band 6 is labeled "Economy Energy" and its size is 13 infinite. When the Northwest surplus is large enough to 14 fall into Band 6, it is due to a combination of favorable 15 hydroelectric conditions, healthy thermal plants, and 16 near-normal or below-normal loads. Under these 17 conditions the market price would generally be set based 18 on the price that it would take to displace the large 19 thermal plants with the lowest incremental operating 20 costs, such as the Colstrip units. The Band 6 price has 21 been set at $8.00/Mwh based on the Colstrip incremental 22 operating cost. 23 Q. What level of loads and resources are used 24 in the Dispatch Model in determining the normalized 25 short-term sales, short-term purchases, and thermal fuel 400 Norwood, Di 12 WWP 1 expenses? 2 A. For retail loads, the Company used the 3 actual 1997 test period net system load by month, 4 adjusted to remove the effects of weather abnormalities, 5 and adjusted to include the DADS direct access program 6 loads that were served by other energy providers during 7 the test period. The test period wholesale contract 8 rights, contract obligations, and energy resources were 9 adjusted to reflect known and measurable changes for the 10 period July 1, 1999 to June 30, 2000, as explained 11 earlier in my testimony. 12 13 / 14 15 / 16 17 / 18 19 20 21 22 23 24 25 401 Norwood, Di 12A WWP 1 Q. What are the dependable capacity values and 2 equivalent availability factors used in the Dispatch 3 Model for the Company's thermal generation? 4 A. The following table shows the dependable 5 capacity and the equivalent availability factors (EAF) 6 included in the Dispatch Model for the Company's thermal 7 generation: 8 Dependable Capacity EAF Plant (Megawatts) (%) 9 Colstrip Units 3&4 222 83% Centralia 201 88% 10 Kettle Falls 46 91% Rathdrum 176 84% 11 Northeast 69 85% 12 The dependable capacity number represents the 13 peak capacity from the plant that can be relied upon over 14 time based on recent performance. The annual equivalent 15 availability factor is calculated as: 1) the amount of 16 energy that could be produced if the plant is operated to 17 the fullest extent possible for each hour of the year, 18 divided by 2) the energy that could be produced if the 19 plant were operated at the dependable capacity for each 20 hour of the year. The difference between the EAF 21 percentage and 100% is related to scheduled maintenance, 22 forced outages, and periods when the units are de-rated 23 and will not run at their full capability. 24 The EAF numbers used in the Dispatch Model for 25 Colstrip, Centralia and Kettle Falls are based on the 402 Norwood, Di 13 WWP 1 average EAFs for the respective plants for the last five 2 years. The dependable capacity figures are based on the 3 most recent performance of the plants. 4 Q. Given the inputs to the Dispatch Model that 5 you have described above, please describe how the 6 normalized level of short-term sales, short-term 7 purchases, and thermal fuel expense are calculated by the 8 model. 9 10 / 11 12 / 13 14 / 15 16 17 18 19 20 21 22 23 24 25 403 Norwood, Di 13A WWP 1 A. As illustrated on Exhibit No. 7, for each 2 of the 720 months (60 years x 12 months), the Company's 3 hydro generation and firm contract rights are compared 4 with the Company's retail load and firm wholesale 5 contract obligations to determine a surplus or deficiency 6 prior to running thermal generation. The Dispatch Model 7 then uses the regional surplus to select the appropriate 8 market price for the purchase or sale of short-term 9 energy. Energy deficiencies are met by the lowest cost 10 alternatives including purchases of short-term energy 11 and/or running the Company's thermal generation. 12 Surplus energy from the Company's system is sold at 13 the market price determined by the regional surplus. If 14 the market price is in excess of the incremental cost of 15 the Company's thermal resources, and those thermal 16 resources are not needed to serve retail load or other 17 firm obligations, the model will run the thermal 18 generation and sell it at the market price. 19 After the Dispatch Model has simulated all 60 years 20 of operation, the short-term sales, short-term purchases, 21 and thermal fuel expenses produced for the 60 years are 22 averaged. These averages are used as the normalized 23 level of short-term sales, short-term purchases, and 24 thermal fuel expenses for rate making purposes. 25 Q. Do the market prices in the Dispatch Model 404 Norwood, Di 14 WWP 1 include charges to recover wheeling and losses associated 2 with the wholesale sales? 3 A. No. The market prices in the Dispatch 4 Model reflect the net price at the Company's system for 5 short-term purchases and short-term sales. Normally, 6 when short-term energy is sold the wheeling charges are 7 paid by the delivering party, who in turn adds the cost 8 to the total charges paid by the receiving party. Since 9 the prices used in the model do not reflect wheeling 10 costs, I did not assume a wheeling expense in Account 11 No. 565. Line 44 of Exhibit No. 6 shows the pro forma 12 expense to be zero. 13 14 / 15 16 / 17 18 / 19 20 21 22 23 24 25 405 Norwood, Di 14A WWP 1 IV. DISPATCH MODEL RESULTS 2 Q. What are the results from the dispatch 3 model for short-term sales, short-term purchases, and 4 thermal fuel expenses? 5 A. The pro forma level of short-term sales, 6 short-term purchases, and thermal fuel expenses from the 7 Dispatch Model are summarized on Exhibit No. 9. These 8 figures are included in Exhibit No. 6 as part of the 9 power supply pro forma adjustments. 10 Q. How does the thermal generation from the 11 Dispatch Model compare to actual generation for 1996, 12 1997 and 1998? 13 A. The following table shows the dependable 14 capacity for each of the plants, and compares the actual 15 generation for 1996-1998 to the output from the Dispatch 16 Model: 17 Generation Dependable Dispatch 18 Capacity 1996 1997 1998(1) Model Plant (Mw) (aMw) (aMw) (aMw) (aMw) 19 Colstrip 222 106 140 185 161 Centralia 201 145 117 150 138 20 Kettle Falls 46 32 32 37 31 Rathdrum 176 33 11 38 22 21 Northeast 69 0 0 1 0 22 (1) Includes estimates for November and December 1998. 23 24 Q. How does the average market price from the 25 Dispatch Model compare with the recent actual market 406 Norwood, Di 15 WWP 1 prices experienced by the Company? 2 A. The weighted average market price from the 3 Dispatch Model, as shown on Line 5 of Exhibit No. 9, is 4 $18.32/Mwh. This price represents the weighted average 5 of all short-term sales 6 7 / 8 9 / 10 11 / 12 13 14 15 16 17 18 19 20 21 22 23 24 25 407 Norwood, Di 15A WWP 1 and purchases from the 60-year analysis in the Dispatch 2 Model. The actual market prices experienced by the 3 Company from 1996 to date are as follows: 4 Weighted Average Market Price 5 Period $/Mwh 6 Pro forma $18.32 7 Jan-Oct 1998 $21.78 1997 $14.53 8 1996 $13.25 9 The actual market prices are adjusted for 10 wheeling and losses (net to the Company's system), and 11 represent a weighted average of short-term sales and 12 purchases by the Company for the respective periods. 13 These prices were taken from the Company's Idaho Power 14 Cost Adjustment (PCA) calculations. 15 In comparing these average market prices it is 16 important to note the streamflow conditions that occurred 17 during each of the periods. As stated earlier, 18 streamflow conditions for 1996 and 1997 were among the 19 highest on record. The January through July runoff at 20 The Dalles, which is generally representative of regional 21 streamflow conditions, was 150% of normal for 1997 and 22 132% of normal for 1996. Flows on the Clark Fork River 23 were 169% and 146% of normal for 1997 and 1996, 24 respectively. These very favorable streamflow conditions 25 contributed to lower short-term market prices, which is 408 Norwood, Di 16 WWP 1 reflected in the actual weighted average market prices 2 for 1996 and 1997. The January through July runoff at 3 The Dalles for 1998 was 98% of normal. The Company's 4 hydroelectric generation for 1998 is expected to be 5 slightly below normal. 6 7 / 8 9 / 10 11 / 12 13 14 15 16 17 18 19 20 21 22 23 24 25 409 Norwood, Di 16A WWP 1 I believe the rate of $18.32/Mwh from the Dispatch 2 Model is a reasonable representation of the weighted 3 average market price under normal streamflow conditions, 4 and current market pricing and thermal operating 5 conditions. 6 7 V. WHOLESALE MARKETING AND ENERGY TRADING 8 Q. Please explain the large adjustments to 9 short-term purchases, on Line 1 of Exhibit No. 6, and the 10 short-term sales on Line 85. 11 A. The 1997 test period short-term purchases 12 on Line 1 of Exhibit No. 6 include short-term purchases 13 made periodically to serve retail load and wholesale 14 contract obligations, as well as purchases of energy for 15 speculative purposes. The test period short-term sales 16 on Line 85 include sales of surplus energy from the 17 Company's system, and sales for speculative purposes. 18 Short-term transactions include transactions with a term 19 of one year or less. 20 Speculative transactions include wholesale purchases 21 of power that are made exclusively for resale to other 22 wholesale parties. They also include wholesale sales of 23 power that are covered later with purchases (selling 24 short). These transactions are unrelated to purchases 25 made to serve retail load, as well as sales of surplus 410 Norwood, Di 17 WWP 1 power from the Company's system. 2 During the 1997 test period the Company made 3 short-term purchases of 12,283,000 Mwh (1,402 aMw). The 4 Company made short-term sales of 12,103,000 Mwh (1,382 5 aMw). The Company's actual net system firm load for 1997 6 was 933 aMw. The short-term purchases and sales were 7 approximately 1.5 times the Company's net system load. 8 It should be noted that the short-term purchases of 12.3 9 million Mwh are approximately equal to the short-term 10 sales of 12.1 million Mwh. The majority of these 11 short-term purchases and sales transactions were for 12 speculative purposes. 13 14 / 15 16 / 17 18 / 19 20 21 22 23 24 25 411 Norwood, Di 17A WWP 1 Under normal streamflow conditions, the Company is 2 near load resource balance, i.e., firm resources are 3 approximately equal to firm retail loads and wholesale 4 contract obligations. Line 4 of Exhibit No. 9 shows 5 short-term purchases for the pro forma period of 923,400 6 Mwh (105 aMw) and Line 2 shows short-term sales of 7 493,700 Mwh (56 aMw). On a net basis for the 8 twelve-month period, based on normal hydroelectric 9 conditions, the Company is a net purchaser of 49 aMw 10 (105 aMw - 56 aMw). 11 The Company's base-load thermal plants, however, are 12 not running at their full availability in the Dispatch 13 Model. The thermal plants run at a lower level in the 14 Dispatch Model because purchases of energy from the 15 market is sometimes lower cost than running the thermal 16 plants at their incremental operating costs (economic 17 displacement). If Colstrip, Centralia, and Kettle Falls 18 were run at their equivalent availability factors, it 19 would result in the following load/resource situation for 20 the twelve-month period: 21 22 Dispatch Model Results: AMw 23 Short-term Purchases (105) Short-term Sales 56 24 Net Purchases (49) 25 412 Norwood, Di 18 WWP 1 Additional Generation Based On EAFs 2 Colstrip 23 Centralia 39 3 Kettle Falls 11 Net Surplus Energy 24 4 5 Thus, there is not a need for short-term 6 purchases or short-term sales transactions in the range 7 of 1,400 aMw, as occurred in 1997, related to serving 8 firm retail load and wholesale 9 10 / 11 12 / 13 14 / 15 16 17 18 19 20 21 22 23 24 25 413 Norwood, Di 18A WWP 1 contract obligations. The 105 aMw of short-term 2 purchases and 56 aMw of short-term sales from the 3 Dispatch Model are very small compared with the 1997 test 4 period actuals, and are consistent with the Company's 5 load/resource situation. 6 Q. Are there benefits that are flowed through 7 to retail customers from the Company's marketing and 8 trading operations? 9 A. Yes. The benefits from the sale of surplus 10 power from the Company's system, and the benefits from 11 dispatching the least cost resources to serve retail 12 loads are credited to customers through the operation of 13 the Dispatch Model. In addition, the benefits from all 14 long-term wholesale marketing transactions are also 15 credited to customers through the individual long-term 16 purchase and sale contracts that are listed in Exhibit 17 No. 6. 18 However, short-term speculative transactions have 19 not been included in the pro forma power supply revenues 20 and expenses. These speculative transactions are 21 short-term in nature, one-year or less, and there is 22 significant uncertainty as to the volume of the 23 transactions that will occur and the profitability of the 24 transactions. These transactions are not directly 25 related to serving the retail load obligations of the 414 Norwood, Di 19 WWP 1 Company. They are also over-and-above the sales of 2 surplus power from the Company's system resources. 3 Because there is significant uncertainty surrounding 4 the volume and rewards from the short-term speculative 5 transactions, and they are not directly related to 6 serving retail loads, the Company believes that the risks 7 associated with these transactions should reside with 8 shareholders and should be excluded from the retail rate 9 making process. 10 Q. Are the power supply operations within the 11 Resource Optimization Department of WWP related in any 12 way to the energy trading and marketing operations at 13 Avista Energy? 14 15 / 16 17 / 18 19 / 20 21 22 23 24 25 415 Norwood, Di 19A WWP 1 A. No. Avista Energy was set up as a separate 2 non-regulated subsidiary of the Company, and their 3 operations are independent of the power supply operations 4 in the Resource Optimization Department of the regulated 5 utility. The primary business focus of the Resource 6 Optimization Department is to optimize the value of the 7 available generating resources, including the economic 8 dispatch of least cost resources to serve retail loads 9 and wholesale contract obligations. Although the 10 Resource Optimization Department does engage in some 11 level of speculative transactions, it is not set up as a 12 full-scale energy trading and marketing operation. 13 VI. SUMMARY OF POWER SUPPLY ADJUSTMENTS 14 Q. Please summarize the net effect of the 15 adjustments to the 1997 test period power supply revenues 16 and expenses on Exhibit No. 6. 17 A. The net effect of the adjustments to the 18 1997 test period power supply revenues and expenses is an 19 increase in net expenses of $46,764,000, on a system 20 basis, as shown on Line 130, Column (c), of Exhibit 21 No. 6. The Idaho allocated portion of this adjustment of 22 $15,516,000 is incorporated into the revenue requirement 23 calculation for the Idaho jurisdiction by Witness 24 Falkner. 25 Q. Does that conclude your direct testimony? A. Yes, it does. 416 Norwood, Di 20 WWP 1 I. INTRODUCTION/SUMMARY 2 Q. Please state your name, the name of your 3 employer and your business address. 4 A. My name is Kelly O. Norwood. I am employed 5 by Avista Corporation at 1411 East Mission Avenue, 6 Spokane, Washington. 7 Q. Have you previously filed direct testimony 8 in this proceeding? 9 A. Yes. 10 Q. What is the purpose of your rebuttal 11 testimony? 12 A. My testimony will respond to Mr. Peseau's 13 testimony regarding the number of years of water record 14 to use in the normalization of power supply costs. I 15 will also address Mr. Peseau's testimony regarding the 16 handling of secondary transactions for ratemaking 17 purposes. 18 Q. Please summarize your rebuttal testimony 19 related to the number of water years for normalization of 20 power supply costs. 21 A. Mr. Peseau's proposal related to the use of 22 a 30-year water record for normalization of power supply 23 costs should be rejected for at least the following four 24 reasons: 25 1. In the current case, as well as in previous cases, 417 Norwood, Di-Reb 1 Avista 1 the Company has consistently used the full water 2 record from the regional hydro regulation studies 3 for the normalization of power supply costs. 4 2. Absent a trend in the water record data, use of 5 the maximum amount of reliable data will produce 6 the best estimate. No credible studies have been 7 put forth to show conclusively that there are 8 trends or cycles in the water record. Water 9 record data is not available for the Clark Fork 10 River prior to 1928, therefore the reliable data 11 consists of data from 1928 to date. 12 13 / 14 15 / 16 17 / 18 19 20 21 22 23 24 25 418 Norwood, Di-Reb 1A Avista 1 3. In the last two rate cases for Idaho Power 2 Company, the Commission has approved the use of 63 3 years and 65 years of water record, respectively. 4 Mr. Peseau's testimony regarding a prior 5 Commission decision related to 20-25 years of 6 water record for IPC is outdated and misleading. 7 In addition, irrespective of prior Commission 8 decisions related to IPC, it should not be assumed 9 that any decision related to IPC should 10 automatically be applied to Avista Utilities. The 11 two companies have different hydroelectric 12 resources, on different rivers, that flow from 13 different drainage basins. There has been no 14 analysis presented in this case that demonstrates 15 that a water record different than the 60 years 16 proposed by the Company would provide a better 17 estimate of normalized power supply costs for 18 Avista Utilities. 19 4. If we were to accept Mr. Peseau's proposition of a 20 30-year water cycle, use of the 60-year record 21 proposed by the Company would include exactly two 22 cycles. Use of two cycles would provide a better 23 estimate than the use of only one cycle. 24 25 Q. Please summarize your rebuttal testimony 419 Norwood, Di-Reb 2 Avista 1 regarding the handling of secondary transactions for 2 ratemaking purposes. 3 A. Mr. Peseau's proposal to include 90% of the 4 secondary sales and purchases as a proforma adjustment 5 should be rejected. As I will explain later in my 6 testimony, the net difference between short-term sales 7 and short-term purchases cannot be used as a measure of 8 profit. This net difference is driven primarily by the 9 Company's load and resource balance. Adjustments to the 10 components that make up this load and resource balance 11 have already been included in the power supply proforma 12 adjustments originally submitted by the Company. 13 14 / 15 16 / 17 18 / 19 20 21 22 23 24 25 420 Norwood, Di-Reb 2A Avista 1 Mr. Peseau's proposal to allocate $3.9 million 2 (Idaho jurisdictional share) of corporate A&G and 3 overhead to speculative trading activities should be 4 rejected. The costs associated with these activities do 5 not even begin to approach the costs that Mr. Peseau 6 proposes to allocate to them. Mr. Peseau's proposal to 7 use energy to allocate costs to the commercial trading 8 activities is inappropriate and is inconsistent with the 9 very important ratemaking principle of "cost causation." 10 The costs assigned to the commercial trading activities 11 should be representative of the costs necessary to 12 support the activity. I have included an analysis later 13 in my rebuttal testimony to estimate the A&G and 14 corporate overheads associated with speculative trading 15 operations. The analysis shows that any adjustment in 16 this proceeding to allocate costs to these trading 17 activities should be no more than $471,000 on a system 18 basis, which is equal to $157,200 for the Idaho 19 jurisdiction. 20 21 II. HISTORICAL WATER RECORD FOR NORMALIZATION 22 23 Q. On Page 24 of his testimony, Mr. Peseau 24 recommends the use of a 30-year water record for the 25 normalization of power supply expenses. Do you have any 421 Norwood, Di-Reb 3 Avista 1 comments on this recommendation? 2 A. Yes. In support of his recommendation, 3 Mr. Peseau has made a number of statements in his 4 testimony that are misleading. For example, on Page 20 5 of his testimony he states: 6 "I take exception to the means by which Avista computes this simple average of annual power 7 costs, as it happens to produce the highest possible figure for test year net power supply 8 expenses." (underscore added) 9 10 / 11 12 / 13 14 / 15 16 17 18 19 20 21 22 23 24 25 422 Norwood, Di-Reb 3A Avista 1 This statement is simply not true. On Page 1 of 2 Exhibit No. 23 I used the same methodology to determine 3 power costs that Mr. Peseau did on his Exhibit No. 206, 4 i.e., the average of purchases and sales for resale for 5 the respective periods. As can be seen on Page 1 of 6 Exhibit No. 23, there are many combinations of the same 7 water record that would produce a higher level of power 8 supply expenses than that proposed by the Company, some 9 of them significantly higher. 10 Although in this case the Company could have 11 developed a case to support some other combination of 12 water record to support higher power supply costs, it did 13 not. The water record proposed by the Company in this 14 case is consistent with that used by the Company in prior 15 rate cases, i.e., the full record of hydroelectric data 16 included in the regional studies, as explained in my 17 direct testimony beginning on Page 8. Thus, Mr. Peseau's 18 statement that the Company used a water record that would 19 produce the "highest possible figure" is untrue and is 20 misleading. 21 22 "Selectively Choosing" Water Record Time Periods for 23 Normalization 24 On Page 20 of his testimony, with regard to the 25 Company's proposed 60-year average of power supply costs, 423 Norwood, Di-Reb 4 Avista 1 Mr. Peseau states: 2 "The average computed within the power cost model is sensitive to the time period chosen and the 3 number of years included in the average. By selectively choosing these time periods and the 4 number of years in the average, one can raise or lower test year power cost estimates." (underscore 5 added) 6 On Page 21 Mr. Peseau refers to "Avista's choice 7 of the sixty year period ending in 1988." 8 As I stated earlier, the Company has consistently 9 used the full water record from the regional hydro 10 regulation studies to normalize power supply costs, and 11 has not engaged in the 12 13 / 14 15 / 16 17 / 18 19 20 21 22 23 24 25 424 Norwood, Di-Reb 4A Avista 1 practice of "selectively choosing" water records to 2 manipulate normalized power supply costs. The Company is 3 fully aware that the normalized power supply costs can be 4 manipulated by "selectively choosing" periods of time 5 from the water record. The Company, however, has chosen 6 not to do so, and has included all of the data available 7 from the Northwest Power Pool regional study. 8 In fact, it is Potlatch, through Mr. Peseau, who has 9 selected a period of water years that would produce a 10 result that would be advantageous to Potlatch. In 11 addition, the study that Mr. Peseau attached as Exhibit 12 No. 207 to support his recommendation of 30 years does 13 not even mention 30 years with regard to cycles. This 14 study is an unpublished study that Mr. Peseau appears to 15 have obtained off of the Internet. Furthermore, on the 16 first page of the study, with regard to using the data to 17 predict the next cycle, it states that "It cannot be 18 demonstrated conclusively that this is so." 19 20 Water Record Approved for Idaho Power Company 21 On Page 23 of his testimony, Mr. Peseau discusses a 22 prior decision by the Commission to use a water record of 23 20-25 years for Idaho Power Company (IPC). Mr. Peseau 24 fails to point out, however, that in the last two rate 25 cases for IPC, 1992 and 1994, the Commission has approved 425 Norwood, Di-Reb 5 Avista 1 the use of 63 years and 65 years, respectively, of water 2 record data to determine power supply costs. 3 Pages 2 to 4 of attached Exhibit No. 23 are pages 4 from IPC Witness Said's testimony in Case No. 5 IPC-E-92-25. Mr. Said states that "63 water conditions" 6 (1928 - 1990) were used for power supply normalization. 7 Page 5 of Exhibit No. 23 is a page from the Commission's 8 order in that case which indicates that the use of 63 9 years was approved for normalization. 10 11 / 12 13 / 14 15 / 16 17 18 19 20 21 22 23 24 25 426 Norwood, Di-Reb 5A Avista 1 Pages 6 and 7 of Exhibit No. 23 are pages from IPC 2 Witness Gale's testimony in Case No. IPC-E-94-5, that 3 indicates that 65 water years (1928 - 1992) were used in 4 the case. Pages 8 to 10 of Exhibit No. 23 are pages from 5 IPUC Staff Witness Hessing's testimony in the same 6 proceeding, that also includes references to the water 7 record from 1928 - 1992. The Commission's order in the 8 case contains no reference to any other alternative water 9 record proposed in the case. The use of 65 years of 10 water record was approved as filed by IPC. 11 In fact, as evidenced by the testimony in the cases, 12 and the Commission's orders, the water record issue was 13 not even a contentious issue in either the 1992 case or 14 the 1994 case. Mr. Peseau's testimony, in this case, 15 regarding the use of 20-25 years averages for IPC is 16 outdated and is misleading. 17 In summary, Mr. Peseau's recommendation related to 18 the use of 30 water years should be rejected for at least 19 the following reasons: 20 1. In the current case, as well as in previous cases, 21 the Company has consistently used the full water 22 record from the regional hydro regulation studies 23 for the normalization of power supply costs. The 24 Company has not engaged in the practice of 25 "selectively choosing" water records to 427 Norwood, Di-Reb 6 Avista 1 manipulate normalized power supply costs. 2 2. Absent a trend in the water record data, use of 3 the maximum amount of reliable data will produce 4 the best estimate. No credible studies have been 5 put forth to show conclusively that there are 6 trends or cycles in the water record. Water 7 record data is not available for the Clark Fork 8 River, where the majority of the Company's 9 hydroelectric generation resides, prior to 1928, 10 therefore the reliable data consists of data from 11 1928 to date. 12 3. In the last two rate cases for Idaho Power 13 Company, the Commission has approved the use of 63 14 years and 65 years of water record, respectively. 15 Mr. Peseau's testimony 16 17 / 18 19 / 20 21 / 22 23 24 25 428 Norwood, Di-Reb 6A Avista 1 regarding a prior Commission decision related to 2 20-25 years of water record for IPC is outdated 3 and misleading. In addition, irrespective of 4 prior Commission decisions related to IPC, it 5 should not be assumed that any decision related to 6 IPC should automatically be applied to Avista 7 Utilities. The two companies have different 8 hydroelectric resources, on different rivers, that 9 flow from different drainage basins. The 10 Commission has been careful in the past to avoid 11 using the "cookie cutter" approach for the 12 companies that it regulates, recognizing that 13 there are differences between the companies. 14 There has been no analysis presented in this case 15 that demonstrates that a water record different 16 than the 60 years proposed by the Company would 17 provide a better estimate of normalized power 18 supply costs for Avista Utilities. 19 4. If we were to accept Mr. Peseau's proposition of a 20 30-year water cycle, use of the 60-year record 21 proposed by the Company would include exactly two 22 cycles. Use of two cycles would provide a better 23 estimate than the use of only one cycle. 24 III. TREATMENT OF SECONDARY TRANSACTIONS FOR RATEMAKING PURPOSES 25 429 Norwood, Di-Reb 7 Avista 1 Q. Do you have any general comments regarding 2 Mr. Peseau's testimony related to secondary transactions 3 before getting into the details? 4 A. Yes. After reading Mr. Peseau's testimony, 5 I am concerned that there may be some confusion as to the 6 secondary transactions that have been included by the 7 Company in this proceeding for ratemaking purposes. For 8 example, on Page 9 of his testimony, beginning on Line 1, 9 he states that: 10 11 / 12 13 / 14 15 / 16 17 18 19 20 21 22 23 24 25 430 Norwood, Di-Reb 7A Avista 1 "Secondary purchases are typically designed either to supplement a company's existing 2 resources until load growth is sufficient to justify the addition of a new large baseload 3 plant, or to take advantage of prices that are below the variable operating costs of its own 4 generating plants. Conversely, secondary sales are made primarily to minimize resource surpluses 5 immediately following the construction of new plants or, in the Northwest in particular, to take 6 advantage of surplus hydroelectric generation." 7 "Ordinarily, the costs and benefits of these secondary purchases and sales are passed through 8 to retail ratepayers as an adjustment to jurisdictional revenues and expenses." 9 10 Mr. Peseau's description of secondary purchases that 11 are used to serve retail load, and sales of power that is 12 surplus to the Company's system is generally correct. 13 His use of the word "ordinarily," however, may leave the 14 impression that these types of transactions have not been 15 credited to customers in this case. 16 As I explained on Page 19 of my direct testimony, 17 the benefits from the sale of surplus power from the 18 Company's system, and the benefits from dispatching the 19 least cost resources to serve retail loads have been 20 credited to customers in this case for ratemaking 21 purposes. The Company's filing provides to retail 22 customers the full benefits of all secondary purchase and 23 sales transactions associated with the operation of the 24 Company's power resources, and transactions related to 25 serving retail load. These transactions will be referred 431 Norwood, Di-Reb 8 Avista 1 to in the remainder of my rebuttal testimony as "system 2 secondary transactions." 3 The only secondary transactions that are not 4 included in the Company's filing for ratemaking purposes 5 are the speculative or commercial short-term transactions 6 that are unrelated to operation of the Company's 7 resources or serving retail load. These speculative 8 transactions will be referred to in the remainder of my 9 rebuttal testimony as "commercial secondary 10 transactions." 11 Q. On Page 11 of his testimony Mr. Peseau 12 states that "short term sales are not inherently 13 speculative, as the Company's testimony suggests." He 14 then goes on to say that 15 16 / 17 18 / 19 20 / 21 22 23 24 25 432 Norwood, Di-Reb 8A Avista 1 "Secondary purchases and sales have almost always 2 included short term transactions that were nevertheless 3 credited to ratepayers." Do you agree with these 4 statements? 5 A. No. These are general statements that 6 appear to encompass all secondary transactions, and 7 Mr. Peseau has failed to distinguish between secondary 8 transactions that are speculative in nature and those 9 that are not. He is partially correct in his statements 10 in that the system secondary transactions are generally 11 not speculative in nature, and have in the past been 12 credited to customers. As stated earlier, these 13 transactions have been fully credited to customers by the 14 Company in this case. 15 The commercial secondary transactions, however, are 16 speculative in nature, and in the fourteen years that I 17 have been involved in the normalization of power supply 18 costs for ratemaking purposes, the Company's commercial 19 secondary transactions have never been included by the 20 Commission for ratemaking purposes. 21 Q. How are the commercial secondary 22 transactions different than the system secondary 23 transactions? 24 A. Commercial secondary transactions entered 25 into by the Company have the following distinguishing 433 Norwood, Di-Reb 9 Avista 1 characteristics: 2 1. Commercial secondary transactions are generally 3 entered into with the full intention of later 4 entering into an offsetting transaction to close 5 out the position, i.e., a purchase of energy is 6 made with the intent to later sell the same amount 7 of energy at a higher price. The Company may also 8 sell energy (sell short) with the intent to later 9 purchase the same amount of energy at a lower 10 price. 11 2. Commercial transactions are not dependent upon or 12 related to the Company's generating resources, 13 i.e., the Company's generating resources are not 14 used to support the transactions. 15 16 / 17 18 / 19 20 / 21 22 23 24 25 434 Norwood, Di-Reb 9A Avista 1 3. The commercial transactions are unrelated to 2 transactions that are used to serve retail load or 3 long-term wholesale obligations. 4 4. Customers' rates are not impacted in either a 5 positive or negative way from the transactions. 6 5. Shareholder capital is placed at risk through the 7 commercial transactions, and the gains and losses 8 are absorbed by shareholders. 9 The commercial trading operations are very similar 10 to an individual with a brokerage account trading stocks 11 in the stock market on a short-term basis. The 12 individual places their own money at risk by choosing to 13 buy shares of stock with the intention of selling the 14 stock later at a higher price. The individual absorbs 15 any gains or losses associated with the transactions. In 16 the same way, the Company places its own money at risk to 17 engage in trading operations. It is appropriate that the 18 Company also absorb the gains and losses associated with 19 this activity. 20 Q. On Page 12 of his testimony, Mr. Peseau 21 asserted that "It is as if a stockbroker traded on a 22 customer's account and then at the end of the year 23 claimed that 90% of the transactions were for the 24 broker's benefit." Is this a proper analogy? 25 A. No, in fact it is just the opposite case. 435 Norwood, Di-Reb 10 Avista 1 Under existing ratemaking, shareholder dollars are at 2 risk, not customer dollars. The Company chooses to place 3 a portion of its earnings at risk for the opportunity to 4 profit from the commercial secondary transactions. The 5 Company's generating resources are not used to support 6 the transactions, and under existing ratemaking, any 7 actual gains and losses from the transactions, as they 8 occur, would not affect customers' rates. Therefore, any 9 gains and losses from commercial transactions should also 10 be excluded from the retail ratemaking process as 11 proposed by the Company, and supported by Commission 12 Staff in this case. 13 Q. Can you provide an example of these 14 commercial secondary transactions? 15 16 / 17 18 / 19 20 / 21 22 23 24 25 436 Norwood, Di-Reb 10A Avista 1 A. Yes. Page 11 of Exhibit No. 23 includes 2 copies of two Deal Tickets for commercial transactions. 3 Deal Ticket No. 677, dated May 4, 1999, is for a sale of 4 25 Mw per hour of firm energy during on-peak hours for 5 the period July 1, 1999 to September 30, 1999 (30,800 6 megawatt-hours or Mwhs), to be delivered to the 7 California-Oregon Border (COB). The energy was sold at 8 $36.75 per Mwh to Statoil. The second Deal Ticket (No. 9 678), dated May 5, 1999, is for a purchase of the same 10 amount of energy for the same period of time, at the same 11 delivery point. The energy was purchased from ECI at 12 $36.60 per Mwh. The names of the individuals executing 13 the transactions have been blacked-out for competitive 14 reasons. 15 Under these transactions the Company sold energy 16 (sold short) with the intention to repurchase a like 17 amount of energy later at a lower cost. In this case the 18 Company sold 30,800 Mwh of energy for $1,131,900 and 19 repurchased 30,800 Mwh of energy for $1,127,280 for a 20 profit of $0.15 per Mwh or $4,620. Although Avista 21 Utilities has an obligation to deliver 25 Mw per hour to 22 Statoil beginning July 1, 1999, it has an equal 23 offsetting right to 25 Mw per hour from ECI for the same 24 period, at the same delivery point. For scheduling 25 purposes, the offsetting transactions are "booked out," 437 Norwood, Di-Reb 11 Avista 1 and the Company has no net obligation. The Company 2 retains the difference between the sales revenue and 3 purchase expense. 4 These transactions also illustrate the relatively 5 high volume of Mwhs and revenue and expense dollars 6 associated with these types of transactions. The margin 7 of $4,620 on this transaction is less than one-half of 8 one percent of the sales revenue. Not only are the 9 margins slim on these transactions, they are not always 10 positive. In some cases the Company must close out the 11 positions at a loss. 12 As I stated earlier, these transactions are not 13 dependent on the Company's resources, they are unrelated 14 to serving retail load, and the Company is at risk for 15 the gains and losses. 16 17 / 18 19 / 20 21 / 22 23 24 25 438 Norwood, Di-Reb 11A Avista 1 Therefore, the Company has excluded the transactions for 2 ratemaking purposes, consistent with previous rate cases. 3 Q. On Page 11 of his testimony, Mr. Peseau 4 asserts that the Company is proposing to change the rules 5 related to secondary transactions. Is the Company 6 proposing to change the ratemaking practices related to 7 secondary transactions in this case? 8 A. No, not at all. As I stated earlier, in 9 all of the fourteen years that I have been involved in 10 the determination of power supply costs for ratemaking 11 purposes, the Company's commercial secondary transactions 12 have never been included by the Commission for ratemaking 13 purposes. These transactions are short-term in nature, 14 generally one to three months in duration, and are not 15 "known and measurable" for ratemaking purposes. The 16 contracts for the commercial transactions that occurred 17 during the 1997 test period have all terminated. Any 18 commercial transactions that may occur during the 19 proforma period generally do not exist at the time of the 20 Company's rate filing and are not known and measurable. 21 On Page 11 of his testimony, Mr. Peseau states that 22 "if a particular transaction relied either in whole or 23 in part on the utility's resources, then the ratepayers 24 have a legitimate claim to at least a portion of the 25 proceeds." These commercial secondary transactions do 439 Norwood, Di-Reb 12 Avista 1 not rely "in whole or in part on the utility's resources" 2 and therefore, by Mr. Peseau's own testimony, the 3 transactions should be excluded from the ratemaking 4 process. 5 Q. On Page 13 of his testimony, Mr. Peseau 6 makes the following assertions regarding the power supply 7 model used to normalize power costs for this case: "The 8 model was built to estimate financial results under very 9 different conditions than those existing today. It is 10 predicated on the old world of regulated wholesale and 11 retail transactions, and it is therefore not a 12 13 / 14 15 / 16 17 / 18 19 20 21 22 23 24 25 440 Norwood, Di-Reb 12A Avista 1 reliable indicator of today's situation of wide open 2 markets with many players." Do you agree with these 3 assertions? 4 A. Absolutely not. First, all wholesale 5 transactions are still regulated by FERC, and retail 6 service is still regulated by the state utility 7 commissions. Although there has been a significant 8 increase in the number of players in the wholesale power 9 market, and the volume of transactions, in general, has 10 increased substantially, these changes have created 11 increased transparency in the market price of power, and 12 have increased the liquidity in the market place, both of 13 which are good for wholesale and retail customers. 14 These changes, however, have not undermined the 15 purpose for which the power supply model was designed. 16 In fact, the power trading companies have built power 17 supply models similar to those relied upon by the Company 18 in this case, to assist in their energy trading and 19 marketing activities. These models are used to forecast, 20 on a short-term basis, the available surplus power in the 21 Northwest region and the market price of power. These 22 models include inputs such as regional load obligations, 23 available hydroelectric and thermal generation, and the 24 price of natural gas. 25 The power supply model still accomplishes today what 441 Norwood, Di-Reb 13 Avista 1 it was designed to do in the past. Inputs to the model 2 have been adjusted to reflect the most recent operation 3 of reservoirs for hydroelectric generation, and the most 4 recent information related to market conditions. The 5 model calculates power supply costs for the period that 6 retail rates will be in effect, based on the retail 7 loads, contract rights, contract obligations, and 8 available thermal generation that will be in place for 9 Avista Utilities during the period, together with 10 normalized streamflow and market conditions. 11 12 / 13 14 / 15 16 / 17 18 19 20 21 22 23 24 25 442 Norwood, Di-Reb 13A Avista 1 Mr. Peseau makes a number specious statements in his 2 testimony regarding the power supply model, apparently to 3 cast doubt about whether the model properly normalizes 4 power supply costs for ratemaking purposes. His 5 statements, however, are not supported by facts in his 6 testimony. He simply makes general statements such as 7 "adjustments of this magnitude are inherently suspect" 8 and "we cannot simply assume that the power supply model 9 is adequate to make adjustments of the size proposed." 10 However, Mr. Peseau identifies no specific shortcomings 11 with the model. 12 The Company's power supply model has been reviewed 13 in detail by Commission Staff in this and in prior rate 14 proceedings. During the discovery phase of this 15 proceeding, the Company spent hours on the telephone with 16 Commission Staff, and in preparing responses to data 17 requests of Commission Staff, regarding the inputs to the 18 power supply model and the operation of the model. 19 Q. On Page 13 of his testimony, Mr. Peseau 20 characterizes the net difference between short-term sales 21 and short-term purchases as a measure of profit. Do you 22 agree with this characterization? 23 A. No. The net difference between short-term 24 sales and short-term purchases cannot be used as a 25 measure of profit. This net difference is driven 443 Norwood, Di-Reb 14 Avista 1 primarily by the Company's load resource balance. For 2 example, if the Company, on average, had resources that 3 exceeded retail load and wholesale obligations, the 4 surplus energy would be sold on the secondary market and 5 you would expect secondary sales revenue to be higher 6 than secondary purchase expenses. The net difference 7 between secondary sales revenue and secondary purchase 8 expense, however, is not a measure of profit, you are 9 simply selling more often than you are buying because you 10 are in a surplus condition. Furthermore, an analysis of 11 "profit" related to secondary sales would include 12 13 / 14 15 / 16 17 / 18 19 20 21 22 23 24 25 444 Norwood, Di-Reb 14A Avista 1 an evaluation of the incremental operating costs 2 associated with any thermal generating units that were 3 run to support the sales. These costs show up in FERC 4 accounts other than Account 555 - Purchased Power 5 Expenses. 6 Lines 1 and 4 of Exhibit No. 9 show the proforma 7 level of secondary sales for this case of 493,700 Mwhs 8 and secondary purchases of 923,400 Mwhs, which makes the 9 Company a net purchaser of secondary energy for the 10 proforma period. This net purchase situation for the 11 Company is based on normalized retail loads, average 12 hydroelectric conditions, firm contract rights and 13 obligations for the proforma period, and the economic 14 dispatch of the Company's thermal resources. Since the 15 Company is expected to make more secondary purchases than 16 secondary sales, the secondary purchase expense would 17 obviously be higher than the secondary sales revenue, as 18 the Company has proposed. This is not an indication, 19 however, that the Company is making uneconomic decisions 20 or incurring "losses," it indicates that the Company, on 21 a net basis, must purchase secondary energy to meet load 22 requirements. 23 It is not surprising that the 1997 secondary sales 24 revenues are slightly higher than the secondary purchase 25 expenses, on a net basis, as Mr. Peseau points out. As I 445 Norwood, Di-Reb 15 Avista 1 explained in my direct testimony, the hydroelectric 2 generation during 1997 was well above normal. The effect 3 of this increased generation, as compared to generation 4 under normal streamflow conditions, was to increase 5 secondary sales and/or decrease secondary purchases. 6 Therefore, on a net basis you would expect to see 7 movement toward a net positive secondary sales revenue 8 figure during a year with favorable hydroelectric 9 conditions. 10 Although the commercial trading activities are 11 included in the secondary sales and secondary purchases 12 figures, because the net difference between the total 13 secondary sales and 14 15 / 16 17 / 18 19 / 20 21 22 23 24 25 446 Norwood, Di-Reb 15A Avista 1 secondary purchases is driven primarily by the Company's 2 load and resource balance, the net difference cannot be 3 used as a measure of profit. 4 Q. On Page 14 on his testimony, Mr. Peseau 5 discusses a possible proforma adjustment to include 90% 6 of the actual 1997 short-term sales and purchases and 7 characterizes such an adjustment as reasonable. Do you 8 agree? 9 A. No, absolutely not. As explained above, 10 the net level of short-term sales and purchases is driven 11 primarily by the Company's load and resource balance and 12 is not a measure of profit. This load and resource 13 balance is affected by changes in retail loads, long-term 14 power contracts, thermal generation, and hydroelectric 15 generation. All of these components are included in the 16 power supply model to determine a normalized level of 17 secondary sales and secondary purchases that are 18 consistent with all other revenues and expenses included 19 in the Company's rate filing. 20 For example, temperatures in 1997 were warmer than 21 normal during the winter months, that resulted in lower 22 retail loads than would occur under normal weather 23 conditions. The lower retail loads resulted in 24 additional energy available at the wholesale level for 25 secondary sales or allowed the Company to make fewer 447 Norwood, Di-Reb 16 Avista 1 secondary purchases to meet the loads, than would 2 otherwise be the case under normal weather conditions. 3 In the Company's rate filing, the Company included 4 an adjustment to reflect higher retail loads, under 5 normal weather conditions, together with an increase in 6 retail revenue associated with the load change. For 7 consistency, on the power supply side the higher retail 8 loads were included in the power supply model to 9 calculate the increased power supply costs associated 10 with serving the higher retail load. The increase in 11 retail load, in the normalization adjustment, causes an 12 increase in secondary purchases or a decrease in 13 secondary sales, because more energy is 14 15 / 16 17 / 18 19 / 20 21 22 23 24 25 448 Norwood, Di-Reb 16A Avista 1 required to serve retail loads. Thus, the normalization 2 adjustment for retail loads has a direct affect on the 3 load and resource balance, the normalized level of 4 secondary sales and secondary purchases, as well as 5 retail revenue. 6 Because the net difference between secondary sales 7 revenues and secondary purchase expense cannot be used as 8 a measure of profit, and adjustments to the components 9 that make up the Company's load and resource balance have 10 already been included in the power supply proforma 11 adjustments, the calculation of a percentage of the 1997 12 actual test period secondary sales and purchases is a 13 meaningless calculation. It would be inappropriate to 14 include such a calculation as a proforma adjustment. 15 Q. On Pages 15 and 16 of his testimony, 16 Mr. Peseau recommends an allocation of a portion of 17 corporate overhead and A&G expenses, as well as general 18 plant rate base, to the commercial transactions. Do you 19 have any comments on this recommendation? 20 A. Yes. Mr. Peseau's recommendation is 21 apparently based on the assumption that the commercial 22 transactions are excluded from ratemaking, and that there 23 should be an allocation of utility costs to the 24 commercial trading activities. We would agree that the 25 commercial transactions should be excluded from 449 Norwood, Di-Reb 17 Avista 1 ratemaking, and that there are some incremental A&G costs 2 associated with these commercial transactions. The costs 3 associated with these activities, however, do not even 4 begin to approach the costs that Mr. Peseau proposes to 5 allocate to them. Mr. Peseau has proposed to allocate 6 $3.9 million (Idaho jurisdictional share) to commercial 7 trading, which is equal to approximately $12.0 million on 8 a system basis. 9 To demonstrate the unreasonableness of the proposal, 10 the single commercial secondary sales transaction of 11 30,800 Mwhs, that I described earlier in this testimony 12 (Page 11 of Exhibit No. 23), would receive an allocation 13 of over $30,500 of corporate overhead and A&G costs 14 15 / 16 17 / 18 19 / 20 21 22 23 24 25 450 Norwood, Di-Reb 17A Avista 1 under Mr. Peseau's methodology ($12.0 million/1,382 aMw * 2 30,800 Mwh). The profit on the transaction was only 3 $4,620. The proposed allocation of corporate overhead, 4 A&G and general plant costs (hereafter referred to as A&G 5 and overhead costs) by Mr. Peseau would be over six times 6 the profit on the transaction. The Company obviously 7 could not afford to engage in these transactions if such 8 an allocation were to be made, or if this was a true 9 representation of real costs. 10 The commercial secondary transactions generally 11 involve high volumes of energy and thin margins, as is 12 evident from the example transaction referenced above. 13 The transactions take only a minimal amount of time to 14 execute and process, and because the trading activities 15 of the Resource Optimization Department operate within 16 relatively narrow risk management trading limits, as 17 compared to full-scale trading operations, they require a 18 relatively minor amount of administrative support. Mr. 19 Peseau's proposal to use energy to allocate costs to the 20 commercial trading activities is inappropriate and is 21 inconsistent with the very important ratemaking principle 22 of "cost causation." The costs assigned to the 23 commercial trading activities, therefore, should be 24 representative of the costs necessary to support the 25 activity. 451 Norwood, Di-Reb 18 Avista 1 Q. Has the Company prepared an estimate of the 2 A&G and overhead costs to support the commercial trading 3 activities? 4 A. Yes. Page 12 of Exhibit No. 23 includes an 5 organization chart for the Electric portion of the 6 Resource Optimization Department. Names of individuals 7 have been omitted for competitive reasons. In reviewing 8 the organization chart, we estimate that approximately 9 four net positions could possibly be eliminated if the 10 short-term trading activity of the Company did not exist. 11 If the short-term trading activity did not exist, the 12 Company would still need the 13 14 / 15 16 / 17 18 / 19 20 21 22 23 24 25 452 Norwood, Di-Reb 18A Avista 1 following positions, as shown on Page 12 of Exhibit No. 2 23, to manage contracts and the generating system to 3 serve retail load and wholesale obligations: 4 1. Long-term marketing personnel to manage existing 5 long-term power contracts and execute new ones to 6 serve retail load. 7 2. Power schedulers to pre-schedule power on a 8 short-term basis to serve retail load and 9 wholesale obligations. 10 3. Real-time schedulers 24 hours per day to manage 11 resources to match load on a minute-to-minute and 12 hour-to-hour basis. 13 4. Hydro engineers to manage the hydroelectric system 14 efficiently. 15 5. Staff to manage the Company's interests in the 16 Colstrip and Centralia jointly-owned coal plants. 17 6. Staff to manage and oversee contracts for the 18 acquisition of coal and wood-waste fuel. 19 7. Accountants to handle accounting and billing, risk 20 management to manage risks such as credit risk, 21 and financial risks on open positions, as well as 22 other support personnel. 23 24 In fact, in reviewing the organization chart there 25 is only one position, a trader position, that is 453 Norwood, Di-Reb 19 Avista 1 dedicated primarily to trading activity. All other 2 positions would be necessary to manage the Company's 3 system to serve load. We recognize, however, that the 4 trading activity is supported by other positions and that 5 each position should be analyzed to estimate the amount 6 of time that should be allocated between retail 7 operations and commercial trading activity. 8 The percentages shown on Page 12 of Exhibit No. 23 9 are estimates of the time for each position that could be 10 allocated to commercial trading activity. These 11 estimates were determined based on discussions with the 12 Manager of Resource Optimization, the Director of Risk 13 14 / 15 16 / 17 18 / 19 20 21 22 23 24 25 454 Norwood, Di-Reb 19A Avista 1 Management, and various other individuals within the 2 Department. The total time related to commercial trading 3 activity is equal to 4.1 positions. 4 To estimate the amount of labor associated with 5 commercial trading, the time for each position was 6 multiplied by the salary for the respective positions, 7 and the total was grossed-up for labor-related taxes and 8 benefits to arrive at loaded labor. This figure is shown 9 on the top of Page 13 of Exhibit No. 23. Page 13 also 10 shows cost estimates for allocated office space, 11 furniture costs, computers, telephones and copier costs. 12 The estimated total annual costs associated with 13 commercial trading is $471,000 on a system basis, which 14 is equal to $157,200 for the Idaho jurisdiction. This 15 estimate corroborates Staff Witness Lobb's statement on 16 Page 12 of his testimony that "Staff continues to believe 17 that the incremental operational cost of the speculative 18 activities is relatively small on an Idaho jurisdictional 19 basis." 20 In summary, any adjustment in this proceeding to 21 allocate costs to these trading activities should be no 22 more than $157,200 on an Idaho jurisdictional basis. 23 Q. Did the Company include an adjustment in 24 its filing to allocate costs to the commercial trading 25 activities? 455 Norwood, Di-Reb 20 Avista 1 A. No. 2 Q. Please explain why. 3 A. As I explained on Page 19 of my direct 4 testimony, 100% of the benefits from all long-term 5 wholesale marketing transactions are credited to 6 customers in this filing. These transactions go beyond 7 managing the Company's resources to serve retail load. 8 They involve structuring long-term deals to either reduce 9 power supply and transmission costs or to create 10 additional margins for the Company. 11 12 / 13 14 / 15 16 / 17 18 19 20 21 22 23 24 25 456 Norwood, Di-Reb 20A Avista 1 An example is the Company's long-term sale of 150 Mw 2 of capacity to Portland General Electric. The revenue 3 under this contract is approximately $18 million per 4 year, and the cost to serve the transaction is a small 5 fraction of this revenue. All of the revenues, and 6 benefits, from this transaction are credited to customers 7 in this case, as shown on Line 91 of Exhibit No. 6. 8 The benefits from these transactions are clearly 9 substantial. Because these transactions go beyond 10 managing the Company's resources to serve retail load, a 11 case can be made that the Company should receive at least 12 a portion of the benefits from these transactions. The 13 Company, however, chose not to pursue a share of these 14 margins in this filing, and also did not propose an 15 adjustment related to the A&G and overhead costs 16 associated with commercial trading. 17 On Pages 14 and 15 of Staff Witness Lobb's testimony 18 he describes his analysis of the benefits to customers 19 from the Company's long-term wholesale transactions. He 20 estimates the benefits to be approximately $5.6 million 21 per year. A reasonable share of Mr. Lobb's estimate of 22 these margins for shareholders would be well above the 23 A&G and overhead costs associated with the commercial 24 trading activity. 25 Q. On Page 16 of his testimony, Mr. Peseau 457 Norwood, Di-Reb 21 Avista 1 recommends that the Commission consider a formal 2 rulemaking related to the allocation of overhead costs to 3 wholesale marketing activities. Do you have any comments 4 on this recommendation? 5 A. Yes. The nature and level of activity 6 related to wholesale marketing and trading will be 7 different for each utility regulated by the Commission. 8 Decisions by the Commission regarding these issues should 9 be utility-specific, and should be handled within the 10 future general rate cases of the respective utilities. 11 Q. Does that conclude your rebuttal testimony? 12 A. Yes. 13 14 / 15 16 / 17 18 / 19 20 21 22 23 24 25 458 Norwood, Di-Reb 21A Avista 1 (The following proceedings were had in 2 open hearing.) 3 MR. MEYER: With that, the witness is 4 available for cross. 5 COMMISSIONER SMITH: Mr. Shurtliff. 6 MR. SHURTLIFF: Thank you. 7 8 CROSS-EXAMINATION 9 10 BY MR. SHURTLIFF: 11 Q Mr. Norwood, at page 8 of your direct 12 testimony, you indicate -- I've got the wrong page. In 13 your direct testimony, you indicate that reliable 14 information as to hydro information exists, I think, to 15 the year 1993. Do I correctly understand your testimony 16 in that regard? 17 A I guess I'd like you to point me to a 18 place, but I can state that we're using the 1928 to 1988 19 record in this case consistent with what we've used in 20 the past and that is to use the data that comes from the 21 regional studies and in this case that's been prepared by 22 the Northwest Power Pool. 23 Q Oh, I apologize, it's page 9 on line 9, 24 starting at line 8, "In the most recent update, completed 25 in 1993," that's not the same thing? 459 CSB REPORTING NORWOOD (X) Wilder, Idaho 83676 Avista 1 A No. For that Northwest Power Pool data, 2 actually, it's the regional hydroelectric data, the 3 information is updated once every 10 years and as they 4 update that information, they update the actual 5 streamflows that occurred in the past for updated 6 irrigation and depletion numbers. It's a pretty 7 substantial study that's prepared and that's what I've 8 referenced here. 9 Q In your rebuttal testimony at page 1, you 10 indicate that reliable data consists of data from 1928 to 11 date. Is it reliable to date or is it reliable to 1988, 12 1993 or some other year? 13 A The information that we've used in this 14 case runs through '88 and the reason we've done that is 15 because that information has been adjusted for irrigation 16 and depletion and all those things and it's been run 17 through the hydro regulation model that was developed and 18 is run by the Northwest Power Pool. Bonneville also runs 19 similar studies using a hydro regulation model that's 20 almost identical to the Northwest Power Pool, and what 21 they do is they run the numbers or the historical 22 streamflows through the existing hydro facilities based 23 on historical streamflows. 24 We actually could use data from 1989 to 25 date, but there have been some changes in the hydro 460 CSB REPORTING NORWOOD (X) Wilder, Idaho 83676 Avista 1 plants, our percentages of the rights from, like, the 2 mid-Columbia projects are different, so we would have to 3 go and make some additional adjustments to the data from 4 1989 to date. It could be done. We've chosen not to. 5 We've in the past used all the data available from 6 Northwest Power Pool and that's what we've used in this 7 case. 8 Q But you could do it if you wanted to with 9 some other adjustments? 10 A Yes, we could, we could use more recent 11 data. 12 Q In that regard, the Clark Fork project came 13 on line with WWP in what year, do you recall, 14 Mr. Norwood? 15 A The Clark Fork projects, I believe, were 16 built in the '50s. 17 Q Would it be fair to say that that's when 18 WWP really became a hydro system rather than relying on 19 purchased power? 20 A Well, I believe there were a number of 21 other projects in the Spokane River that were built 22 before 1920, so we've had hydro generation for many 23 years. 24 Q But wasn't the Clark Fork project the major 25 source after it came on line in the '50s and I believe 461 CSB REPORTING NORWOOD (X) Wilder, Idaho 83676 Avista 1 it's '54? 2 A Noxon Rapids in the Clark Fork, on the 3 Clark Fork, as well as Cabinet are our largest projects, 4 but I'm not sure that you could characterize that as a 5 time when we really became a hydro utility. If you look 6 at the generation we had, especially during the time when 7 Long Lake was coming on line, it was a very large project 8 at the time and if you compare it to the loads we had at 9 the time, I'm sure it represented a pretty large portion 10 of our load requirements. 11 Q And I appreciate that this is a simple -- 12 I'm making what you do, your professional analytical 13 work, overly-simplistic, but the purpose of your work, 14 the purpose in selecting a period of time for water, is 15 so that we can suggest that the history is a harbinger of 16 what we can anticipate in the future; is that not 17 correct? 18 A Well, absent knowing what the future will 19 bring in the way of streamflows, then all we have to work 20 with is the past as far as precipitation and therefore 21 streamflows. 22 Q And so you pick a period of time, 23 Dr. Peseau picks a period of time, somebody else could 24 pick a period of time to suggest that that historical 25 information will be replicated in the future? 462 CSB REPORTING NORWOOD (X) Wilder, Idaho 83676 Avista 1 A We have chosen to use for ratemaking in 2 this case as well as in the past all the available 3 information that's used in the region and that's part of 4 the reason we chose it was because it's done by a third 5 party. It's not something that we took and changed. The 6 data on the Clark Fork River at the White Horse Rapids 7 gauging station is available beginning in 1928, so the 8 data prior to that time is not available for our system 9 and we didn't select a different period. 10 We in the past have consistently used the 11 maximum amount of data available. As I said before, we 12 could use from 1999 [sic] forward. It would require a 13 substantial amount of work to adjust that data, to 14 incorporate that and we have not done that, not because 15 we didn't want to, it's just that it's a lot of work and 16 so we've chosen to use what we have used in the past for 17 normalization of power costs. 18 Q I think you said 1998, you meant to say 19 1988, did you not? 20 A We used the data from '28 to '88 21 currently. I'm sorry if I misspoke. 22 Q And you could have used from '88 to present 23 with the additional work? 24 A We could. 25 Q And do you have a seat-of-the-pants 463 CSB REPORTING NORWOOD (X) Wilder, Idaho 83676 Avista 1 judgment based on your expertise as to what impact the 2 addition of those years would be on the result? 3 A I have not run the numbers to see what the 4 impact of that would be. 5 Q Well, you don't think that they were as dry 6 as the '30s, do you? 7 A I know that '94 was a very dry year, '96 8 was a wet year. There's a combination of wet years and 9 dry years in there. 10 Q Could you not create a model and base your 11 proposal on actual experience since 1954, that is, when 12 the Clark Fork major projects came on line, and come from 13 '54 forward and keep rolling? 14 A You could pick any number of water records 15 to estimate numbers. What we've chosen to do in this 16 case is -- and, again, what we're after is the best 17 estimate of normal power supply costs from our hydro 18 generation and absent a trend or some kind of a cycle in 19 the data, then the best information is going to be the 20 maximum amount of data available in determining that 21 estimate and I've seen no studies that would demonstrate 22 a trend in the data, so, therefore, by using all the 23 data, it will provide the best estimate of normalized 24 power costs. 25 The fact that Cabinet came on line at a 464 CSB REPORTING NORWOOD (X) Wilder, Idaho 83676 Avista 1 certain date shouldn't necessarily be a reason why you 2 would pick that date forward. We have records of 3 streamflows from that drainage basin from 1928 and it's 4 reliable information, it's used in the region and it's 5 appropriate, then, to use that data. 6 Q Well, not every expert or decision maker 7 looking at which period of time ought to be used would 8 agree with you, would they? Some people, Dr. Peseau 9 suggested a 30-year period which you take exception to, 10 but other experts, other persons might have other 11 judgments, might they not? 12 A They might and Dr. Peseau mentioned 13 Bonneville and I spoke with Bonneville last week just to 14 verify what they used for ratemaking purposes for the 15 studies that they use and I spoke with Ken Smith who is 16 the manager of the regional coordination there and he's 17 responsible for oversight of the hydro regulation studies 18 done by Bonneville there in Portland, for ratemaking 19 purposes, they use a 50-year study from 1928 to 1978. 20 For their white book study which is used 21 for some of the provisions in their power sales 22 contracts, they also use a 50-year study from 1928 to 23 1978. For what if studies, for operations, they use the 24 60-year period as we propose in this case from 1928 to 25 1988, so they have in the past consistently used the 465 CSB REPORTING NORWOOD (X) Wilder, Idaho 83676 Avista 1 regional data from 1928 forward and we believe that that 2 will provide the best estimate. 3 MR. SHURTLIFF: I have no further 4 questions. 5 COMMISSIONER SMITH: Thank you, 6 Mr. Shurtliff. Mr. Ward. 7 MR. WARD: Madam Chair, I'm going to be 8 some time with Mr. Norwood. I've got a couple of 9 exhibits to pass out if you would like to take a break 10 before we start. 11 COMMISSIONER SMITH: Sure, let's take 10 12 minutes. 13 (Recess.) 14 COMMISSIONER SMITH: We're ready for 15 Mr. Ward. 16 MR. WARD: Thank you. 17 18 CROSS-EXAMINATION 19 20 BY MR. WARD: 21 Q I think what I'd like to do in the way of 22 starting, Mr. Norwood, is to, with your assistance, help 23 explain to the Commission what they may already realize 24 and that is the importance of the power supply adjustment 25 in this proceeding, so with that, I'd like to start by 466 CSB REPORTING NORWOOD (X) Wilder, Idaho 83676 Avista 1 asking you if you'll turn to your direct testimony, 2 page 2, lines 16 through 18. 3 A I'm there. 4 Q And there you say the net effect of PSA -- 5 sorry, the net effect of the power supply adjustment on 6 an Idaho basis is an increase of 15,516,000 in revenue 7 requirement. Do you see that? 8 A Yes, I do. 9 Q And do you recall what the total increase 10 the Company is requesting is in this case? 11 A It's approximately $14 million. 12 Q Okay; so the power supply adjustment 13 exceeds the claimed revenue deficiency? 14 A That's correct. 15 Q And I placed on your table and in front of 16 the Commission four sheets and I apologize for the -- 17 A I don't have copies. 18 MR. WARD: I'm sorry, I'll get you one. 19 (Mr. Ward approached the witness.) 20 Q BY MR. WARD: Now, having handed you that 21 document which consists of four pages, let me represent 22 to you to skip forward a little bit, Mr. Norwood, that 23 these are selected pages out of Exhibit 11 of 24 Mr. Falkner's and I don't intend to ask you to 25 corroborate any of this data, I just want to quickly make 467 CSB REPORTING NORWOOD (X) Wilder, Idaho 83676 Avista 1 a point. If you look down -- and let me further tell you 2 that on the right-hand side of the first page is a column 3 labeled "Restated Total," do you see that? 4 A I do. 5 Q And let me tell you that according to 6 Mr. Falkner, this is the Commission basis results for 7 1997 with a couple of exceptions we're going to discuss. 8 Now, if you look down at the lower right-hand corner, do 9 you see that number that didn't reproduce very well under 10 rate of return? 11 A I'm really not familiar with these sheets. 12 I can follow what you're pointing to, but I'm not 13 familiar with this exhibit. 14 Q All right. The rate of return indicated 15 there is 9.65 percent, is it not? 16 A I can see that, yes. 17 Q Now, let's go to the next sheet and what I 18 want to you look at, focus on, on the next sheet is line 19 29 labeled "Net Operating Income" and there are five 20 columns in that, on that sheet. Do you see that? 21 A I do. 22 Q Now, the net operating income under 23 column 1 entitled "Pro Forma Power Supply" is minus 24 $9,918,000; correct? 25 A I see that, yes. 468 CSB REPORTING NORWOOD (X) Wilder, Idaho 83676 Avista 1 Q And then in the succeeding columns we have 2 a number of pro forma adjustments. Do you see those? 3 A Yes. 4 Q And over on the next page we have three 5 more pro forma adjustments. Do you follow me so far? 6 A Yes. 7 Q Now, of that $9,918,000, that's a number 8 you furnished to Mr. Falkner, is it not, for the power 9 supply adjustment? 10 A Actually, no. That's the net operating 11 income impact. The number I would have provided would 12 have been a little bit different. He's done some work to 13 the numbers that I provided to him. The numbers that I 14 provided to Mr. Falkner would be on my Exhibit No. 6 and 15 where he goes from there I'm not sure. 16 Q And your Exhibit No. 6 -- 17 A Page 3. 18 Q All right. 19 A At the bottom, line 130. 20 Q All right, I have it. 21 A What I provided to him was an adjustment of 22 46,764,000 on a system basis. 23 Q I see. Then did you include in that system 24 basis a gross-up for tax effect? 25 A No, there are no taxes included in the 469 CSB REPORTING NORWOOD (X) Wilder, Idaho 83676 Avista 1 numbers, no income taxes included in the numbers, that I 2 have included on my exhibit. 3 Q If you know, how did Mr. Falkner get to the 4 $9 million figure that appears in his exhibit? 5 A I think you would need to talk to 6 Mr. Falkner about that. 7 Q All right. Let's go back to your direct 8 testimony. If you would turn to page 3 of that 9 testimony -- 10 A I'm there. 11 Q -- here you begin a discussion that runs 12 over into the next pages regarding the power supply 13 adjustment and how it works and I want to ask you some 14 questions about it and see if my understanding is 15 correct. First of all -- and you'll forgive me if I 16 refer to what you call the dispatch model as the power 17 supply model from time to time. 18 A That's fine. 19 Q As I understand it, first of all, the power 20 supply model normalizes for weather, streamflows and 21 prevailing market conditions; is that correct? 22 A The weather component is really related to 23 the retail loads. Those retail loads are corrected for 24 weather. The hydro generation is corrected for 25 streamflows and the market prices that are included are 470 CSB REPORTING NORWOOD (X) Wilder, Idaho 83676 Avista 1 adjusted for the market conditions. 2 Q All right, and the latter adjustment, that 3 is, the market conditions or market prices, is in part at 4 least an external input into the model? 5 A The modeled price, that's correct. It's an 6 external input based on data that I've outlined here in 7 my testimony. 8 Q Yes, I understand that. Now, we'll talk 9 about that part of the model later, but would it be fair 10 to say that there's a second part of the model or of the 11 power supply adjustment and you provide some examples on 12 the bottom of page 3 and throughout page 4 that consist 13 of known contract changes and other items, but I want to 14 talk about the contract changes for the moment. 15 A Yes, I have listed a number of items on 16 pages 3 and 4 related to changes in power costs. 17 Q All right, and I take it that these 18 adjustments, if you turn back to page 3, lines 10 and 11, 19 run through and including the period July 1, 1999, and 20 ending June 30, 2000; is that correct? 21 A The adjustments that I've identified here 22 reflect the changes in power costs from the test period 23 of 1997 to the pro forma period beginning July 1, 1999, 24 ending in June of 2000. 25 Q All right. Now, you would have filed your 471 CSB REPORTING NORWOOD (X) Wilder, Idaho 83676 Avista 1 testimony in December of 1998; is that correct? 2 A I believe that's correct. 3 Q So to the extent we are focusing on a 1999 4 through 2000 year for power supply adjustments, isn't it 5 fair to say that the power supply adjustment test year is 6 July of 1999 through June of 2000? 7 A I guess we're getting balled up here a 8 little bit in semantics. I typically refer to a test 9 year as the historical period where you have the actual 10 numbers. In this case it's 1997. I normally use 11 pro forma period as the period where we're going out to 12 pick up known and measurable changes, which in this case 13 would be July, '99 through June, 2000. 14 Q All right, and those pro forma changes, 15 would it be fair to say that in December of 1998, you 16 could not necessarily know all the pro forma changes that 17 would occur through mid '99 and mid 2000? 18 A The purpose of a normalization adjustment 19 is to reflect those that are known and that are 20 measurable and that's what we've filed here in this case 21 is to look at specific contracts that have provisions 22 that show that the rate will change over time. To the 23 extent that the contract has those known changes in 24 there, we reflected those in this rate case and for the 25 adjustments that are included here, we tried to limit 472 CSB REPORTING NORWOOD (X) Wilder, Idaho 83676 Avista 1 those adjustments to just those ones that are known and 2 measurable. 3 Q I understand that. Do you know whether 4 this Commission has traditionally allowed a full 5 projected test year? 6 A I don't know. 7 Q Let's go to the contracts themselves and 8 see if we can explain how this works. When a contract 9 terminates, would it be correct to say that what happens 10 is the model drops any available power that's displaced, 11 that is available as a result of, let us say, a sale 12 contract terminating, the model drops any available power 13 into the short-term pricing category? 14 A To the extent that a firm contract 15 terminated and there wasn't another one, yes, that's 16 true, it would end up being sold short term. 17 Q I think you said yes, it drops into the 18 short-term category? 19 A Yes, it does. 20 Q Now, and so if we were to take the contract 21 on page 4, let's take the PP&L '89 contract at 9 and 10, 22 at lines 9 and 10, I don't know what the opening price of 23 that contract was or what the prevailing price of that 24 was, but it assumedly had some sort of fixed price, did 25 it not? 473 CSB REPORTING NORWOOD (X) Wilder, Idaho 83676 Avista 1 A Yes, it did. It had a capacity rate as 2 well as an energy rate. 3 Q And so if nothing else happens, when that 4 contract terminates, that, let me call it, load or power 5 will in fact be repriced by the model at short-term 6 rates? 7 A What really happens is that even though the 8 contract terminates and drops out for the pro forma 9 period, the fact that the contract is gone would really 10 be blended with all the other loads and obligations for 11 the Company, and the fact that that contract is gone may 12 provide power to serve increases in load that have 13 occurred over time, and that's really what has occurred 14 for the Company who made a number of long-term contracts 15 many years ago and the PP&L '89 contract was one of 16 those. Some of those contracts are phasing out now. 17 Loads have grown and to some extent the loads that went 18 for these contracts are now being used to serve load. 19 Q I understand that, but load is only growing 20 at roughly two percent a year on Avista's system, is that 21 not correct, retail load? 22 A I heard that number mentioned this morning. 23 Q All right; so if that's a sizeable 24 contract, load growth is not going to pick it up; isn't 25 that true? 474 CSB REPORTING NORWOOD (X) Wilder, Idaho 83676 Avista 1 A Regardless of load growth, what we look at 2 and what's included in the dispatch model are the loads 3 that are available as well as all the known contracts for 4 the pro forma period and, as you indicated, to the extent 5 that you are surplus, it falls into the short-term market 6 and to the extent we're deficient, then we'll buy that 7 from the short-term market. 8 Q Let me ask you, and I realize that I'm 9 going to have to oversimplify here, but what is the 10 average short-term price that we are using for the model? 11 A The weighted average price is $18.3 per 12 megawatt-hour in this case. 13 Q And so if the PP&L contract, for instance, 14 was at $25.00 a megawatt-hour, what happens in terms of 15 effect on revenues, bottom line, is that power gets 16 repriced from, let us say, 25 hypothetically to 18.3? 17 A That would be the case. 18 Q Now, in the real world, notwithstanding the 19 model does that -- well, first of all, let me ask you 20 this: When we have terminating contracts for off-system 21 sales, if load growth is not sufficient to consume that 22 power supply, the likelihood is that the Company will 23 again try to market that particular supply of power, will 24 it not? 25 A To the extent the Company does have 475 CSB REPORTING NORWOOD (X) Wilder, Idaho 83676 Avista 1 long-term surpluses, it certainly would try to sell that 2 power on a longer-term basis to achieve higher margins. 3 Q All right, and so in the real world, it 4 isn't necessarily going to be true that the terminated 5 contract brings only the short-term market price; isn't 6 that also true? 7 A And that really gets to what I said before 8 is for ratemaking purposes, you need to take a look at 9 all the existing contracts, whether it's a purchase or a 10 sale, stack them all up and determine whether you're long 11 or short, need to purchase or buy. In the example 12 earlier where you mentioned the $25.00 versus $18.00 per 13 megawatt-hour, in that particular case, the revenue would 14 go from $25.00 to 18, but that's assuming you're just 15 isolating that one agreement and for ratemaking, we don't 16 isolate one particular agreement, we put them all 17 together so that we do have a total picture of what our 18 power supply costs are. 19 Q I understand that, Mr. Norwood, but you 20 maybe anticipated in a way the point I'm trying to make. 21 When we are preparing testimony in December of '98 and we 22 have terminating sales agreements, the model drops them 23 into the short-term price category, that we can agree on, 24 recognizing your caveat of everything gets mixed in the 25 model and the entire picture is analyzed, but all other 476 CSB REPORTING NORWOOD (X) Wilder, Idaho 83676 Avista 1 things being equal, it will drop them into the short-term 2 price category. 3 Now, the difficulty I have with that or the 4 question I have for you is in the real world when you're 5 sitting there with a contract terminating, let us say, 6 shortly after you're going to file your testimony in 7 December of '98 and you adjust that contract out, you may 8 not know what the replacement sale might be in '99 9 through 2000; isn't that also true? 10 A There will no doubt be changes in power 11 contract costs following the filing, that is true. What 12 we do reflect in the rate case is a point in time those 13 changes that are known and measurable. 14 Q And the flip side of that also occurs on 15 purchases, does it not; that is, if the purchase contract 16 terminates, then all other things being equal, that 17 purchase will go to market price, short-term market 18 price? 19 A Yes. 20 Q But in fact, of course, in the real world 21 the utility will try to beat that short-term market price 22 if it can with a contract, will it not? 23 A In the real world what we do is take a look 24 at our need for resources and if there's a need for 25 long-term purchases, then we'll go out and make that 477 CSB REPORTING NORWOOD (X) Wilder, Idaho 83676 Avista 1 long-term purchase. 2 Q And to cut to the bottom line, Mr. Norwood, 3 with respect to this portion of the power supply model, 4 isn't it subject to exactly the infirmities that argue 5 against the use of a projected test year; that is, the 6 typical argument is that the utility will have every 7 incentive to have a very good forecast of the increased 8 expenses and decreased revenues, but a somewhat less 9 accurate one of the possibility for the reverse of those 10 items? 11 A No, I don't think so at all. In this case 12 the dispatch model incorporates all the long-term 13 contracts, plus and minus, purchases and sales, and if 14 you look at our load/resource balance, I think it's very 15 clear that we're, as mentioned earlier today, pretty 16 close to load/resource balance and what's been reflected 17 in the case are the known and measurable changes. I 18 don't view that as at all looking at a projected test 19 period. 20 The contracts that are included are those 21 contracts that are known to exist beginning July 1 of 22 '99, which is next month, and running through the next 23 12-month period, so the contracts that will be built into 24 rates at the time the order is written will be those 25 contracts that will be in place at the time that rates 478 CSB REPORTING NORWOOD (X) Wilder, Idaho 83676 Avista 1 are set which is entirely appropriate. 2 Q But if in fact you have a contract that 3 terminates, that has a known termination date, let's make 4 it a contract for sale, and you've included it in your, 5 what I'll call, external adjustment to the power supply 6 model, understandably because you know the date that it's 7 going to terminate, let us say February 1, 1999, do you 8 have that in mind? 9 A Yes. 10 Q And at the point where you're preparing 11 your testimony in December of 1998, you don't know that 12 there's going to be any replacement for that and so what 13 the power supply model does is drop it into the 14 short-term market pricing category, again, with the 15 caveats you've suggested -- 16 A Yes. 17 Q -- at 18 point mills, but now let us assume 18 that that was a 25 mill contract and in fact I, your 19 sales representative, am now making as we sit here today 20 a deal to sell that power that was on contract to someone 21 else at 25 mills again. Doesn't it -- isn't it a 22 mathematical truism that if that should happen your power 23 supply adjustment understates the revenue that you will 24 receive? 25 A Those contracts can go either way. That 479 CSB REPORTING NORWOOD (X) Wilder, Idaho 83676 Avista 1 particular case could occur. The opposite case could 2 also occur where there was a need for power and we were 3 to go out and buy power at $25.00 a megawatt-hour, the 4 average price built into the rate case was $18.00 per 5 megawatt-hour and, therefore, we'd end up incurring costs 6 that were not being recovered in rates, so either case 7 could occur. 8 Q But that's a somewhat different case, is it 9 not, Mr. Norwood, that's a case of the market moving on 10 you or your power supply requirements moving on you, 11 which is an item we're trying to normalize with the power 12 supply model; isn't that true? 13 A It wouldn't necessarily be the case of the 14 market moving. It's a matter of the need for, in this 15 case, long-term resources is what I'm assuming if you 16 looked at a 2- to 3-, 5-year contract that may come in or 17 drop out. 18 Q Let me ask it another way: With respect to 19 purchases, isn't it generally true with respect to sales, 20 isn't it generally true that firm contract rates will 21 bring a higher price than short-term market rates, all 22 things being equal? 23 A Yes. 24 Q All right; so if a contract terminates and, 25 yes, we're talking about a hypothetical circumstance, but 480 CSB REPORTING NORWOOD (X) Wilder, Idaho 83676 Avista 1 if a contract terminates, a firm contract terminates, and 2 you get only the short-term price in the model, there is 3 the possibility, is there not, that that contract could 4 in the actual year we're modeling, '99 through 2000, be 5 replaced by yet another firm contract? 6 A Mr. Ward, that possibility certainly does 7 exist. If you look at the dispatch model and the 8 resources that are run against our loads, as I mentioned 9 in my testimony, we're actually a net purchaser of energy 10 during this '99 through 2000 period, so there isn't this 11 excess power that is available to go out and make a firm 12 sale. We're actually short when you look at the 13 load/resource situation in this particular case. 14 Q Actually, as I recall your testimony, 15 Mr. Norwood, didn't you testify that you're not actually 16 short, that you're only -- that the only reason why you 17 purchase, why your purchases are somewhat out of balance 18 is for economic dispatch purposes? 19 A That is correct. 20 Q Now, I don't want to walk through this 21 whole thing again, so let me just say that isn't it true 22 that something similar happens in the case of 23 transmission revenues? And the best way to maybe ask you 24 about this is to ask you to turn to page 3 of Exhibit 6. 25 A I'm there. 481 CSB REPORTING NORWOOD (X) Wilder, Idaho 83676 Avista 1 Q Well, you're ahead of me. Here we have, I 2 take it, under -- I don't know if this refers to an 3 account number, but under line 107 is 456, is that an 4 account number? 5 A Yes, it is. 6 Q All right, other electric revenue and below 7 that is transmission system and wholesale system. Do you 8 see that? 9 A Yes. 10 Q Now, again, you're adjusting through 11 '99-2000 in these cases as well; is that correct? 12 A That's correct. 13 Q And, again, you have the same question of 14 contracts terminating and we can see some of them listed 15 on that in lines 108 through 120; correct? 16 A Yes. 17 Q Now, the net effect of that is that, again, 18 when you're sitting there in December of '98, at the 19 latest, preparing your testimony, what you see is a 20 decrease in revenues from transmission system and 21 wholesale system from 19,785,000 to 16,345,000. Do you 22 see that? 23 A I do. 24 Q And that's a $3,440,000 increase; correct? 25 A Yes. 482 CSB REPORTING NORWOOD (X) Wilder, Idaho 83676 Avista 1 Q Decrease, I'm sorry. 2 A Decrease. 3 Q But in fact, the transmission system is 4 still sitting there, is it not, with the same capability, 5 all other things being equal, that it had in 1997 when 6 you used the base year? 7 A I don't know how the capabilities and 8 opportunities have changed since that period of time, but 9 I would imagine there are some opportunities out there, 10 yes, that remain. 11 Q And really, that's all I'm trying to point 12 out is that when we actually get to '99-2000, some of 13 those opportunities may in fact come to pass; correct? 14 A There will no doubt be changes through this 15 next year and, like I said before, there can be positives 16 and negatives. When we filed the case, again back in 17 November, December, whenever we filed, the Kettle Falls 18 fuel cost was $5.91 and right now it's over $7.00, so if 19 we were to refile the case today, we would increase the 20 Kettle Falls fuel cost to over $7.00, and that's an 21 example of how that at a point in time we pick the known 22 and measurable changes for all of our costs and we put 23 them in the case, recognizing that the next month or the 24 next six months things will change, we know that. 25 Some costs will go up, some costs will go 483 CSB REPORTING NORWOOD (X) Wilder, Idaho 83676 Avista 1 down, but unless we want to have a rate case every month 2 and determine costs every month, we have to pick a point 3 in time and measure the costs that we know that are there 4 and put them in the rate case and that's the way that 5 rates have been set in the past and this case is no 6 different in that we've been consistent with the way the 7 rates have been set in the past. 8 Q But return to the focus on the question, 9 Mr. Norwood, if you would, regarding let us take the 10 transmission system, isn't it a fact when you adjust for 11 the known and measurable changes you know in 1998, to the 12 extent you have terminating contracts, you'll take them 13 out, but the fact of the matter is when you get to '99 14 through 2000, it's a truism, is it not, that the only way 15 actuality could change is with increased revenues, in 16 other words, use of the system that you did not foresee? 17 A No, I don't believe that's true. There are 18 other contract changes that can cause costs to go up as 19 well as you can have increased revenues from the 20 transmission contracts as well as the power supply 21 contracts. Now, there are some contracts that have 22 specific costs, those will not change, but as I said 23 before, you will have costs that will go up for some 24 contracts, you will have revenues that will go up for 25 others. 484 CSB REPORTING NORWOOD (X) Wilder, Idaho 83676 Avista 1 Q But let us hold for the moment operational 2 costs and those sort of considerations aside. The fact 3 of the matter is when you look out projecting from '98, 4 and let me put it in layman's terms that are clearly not 5 accurate, you look out and you say, I see terminating 6 contracts and my current almost 100 percent use of the 7 transmission system is only going to be 80 percent and 8 you adjust your test year for that, but the fact of the 9 matter is a year later you may be back up to 100 percent 10 again; isn't that true? 11 A The opportunities and the revenues that we 12 might get off the transmission system will be dependent 13 upon the opportunities in the market that's out there. 14 The fact that we have available transmission capacity 15 doesn't mean that parties will take advantage of that, 16 and it would be nice if the only way in between rate 17 cases were for revenues to go up, but, unfortunately, 18 that's not the case. Costs can go up as well as revenues 19 can go up. 20 Q Mr. Norwood, let me ask it another way: 21 Doesn't this whole question raise any issue in your mind 22 about a methodology that chooses a 1997 test year for 23 most costs and most revenues and yet for this particular 24 adjustment, power supply, we are relying totally on 25 projections of '99 through 2000? 485 CSB REPORTING NORWOOD (X) Wilder, Idaho 83676 Avista 1 MR. MEYER: You know, I object. Up to this 2 point I've been patient. Questions in one form or 3 another are the same. They've been asked and answered. 4 The witness has testified repeatedly, first of all, that 5 this does not constitute a projected test period; number 6 two, that it is consistent with prior ratemaking practice 7 to take into account for purposes of this adjustment, the 8 power supply pro forma adjustment, known and measurable 9 changes. 10 Staff, as you'll see when they take the 11 stand, has reviewed this as they have in the past and 12 screened out and examined the known and measurable 13 adjustments. Now, Mr. Ward tried repeatedly by 14 rephrasing essentially the same question to get this 15 witness to admit otherwise. We've covered the ground, so 16 I object. 17 COMMISSIONER SMITH: Mr. Ward. 18 MR. WARD: Really, I'm asking him the 19 summary question in light of the cross-examination we've 20 been through. What I asked briefly was whether he saw 21 any potential problem with using historical data for the 22 test year in all respects except the power supply 23 adjustment, which after all is the entirety of the 24 revenue requirement in this case for '99 and 2000 and 25 that is projected. I don't see how counsel can argue 486 CSB REPORTING NORWOOD (X) Wilder, Idaho 83676 Avista 1 that there's any other way to characterize that. We are 2 sitting here in June of '99. 3 COMMISSIONER SMITH: Mr. Meyer, I'm going 4 to overrule your objection. I did see that as the ending 5 question in a line of cross. 6 Mr. Norwood. 7 THE WITNESS: Could I ask him to repeat the 8 question, please? 9 Q BY MR. WARD: Do you see -- let me put it 10 in the negative. Don't you see that there is a potential 11 problem regarding the accuracy of a test year that's made 12 up in part of normal '97 test year data adjusted for 13 known and measurable changes and power supply data, power 14 supply adjustment, that is projected through a '99-2000 15 time frame? 16 A And, again, not at all. I see it's very 17 consistent to include in the case those known contract 18 changes for ratemaking purposes. 19 Q Well, I'll live with that. 20 A One more item, too, with regard to the 21 statement about the power supply adjustment constituting 22 the full rate request. There are a number of other costs 23 that are testified to by other witnesses. Some increase, 24 some decrease, the fact that the power supply adjustment 25 is roughly equal to the increase doesn't mean there are 487 CSB REPORTING NORWOOD (X) Wilder, Idaho 83676 Avista 1 not other costs that have gone up because they certainly 2 have and other witnesses have spoken to those and there 3 obviously have been some decreasing costs or increased 4 revenues that have also been included in the case. 5 Q Well, Mr. Norwood, I have to follow that 6 up. You've really wandered into Mr. Falkner's area, but 7 the truth of the matter is it's a mathematical fact, is 8 it not, that the power supply adjustment exceeds the 9 increased revenue requirement requested in this case? 10 A This adjustment in isolation, it does, 11 that's correct. 12 Q So it's true enough that other things go up 13 and down, but the entirety of this case really has to 14 stand and fall on the power supply adjustment. If it 15 should prove to be invalid, then the Company even if it 16 wins every other issue has no rate case. 17 MR. MEYER: Well, I object. That 18 mischaracterizes the testimony that has been given. This 19 case is about a number of issues. We've had a cost of 20 money expert up there. The power supply adjustment is 21 one component. Now, there are offsetting items that move 22 the other direction, but there are a number of issues. 23 Mr. Ward has been around the ratemaking process long 24 enough to know that. 25 COMMISSIONER SMITH: I don't think Mr. Ward 488 CSB REPORTING NORWOOD (X) Wilder, Idaho 83676 Avista 1 really expected an answer and I think the Commission can 2 do the math. 3 MR. WARD: Thank you, let's move on. 4 Q BY MR. WARD: Now, let's turn to your 5 rebuttal, if you can, Mr. Norwood. Now, I want to return 6 to the subject of the model itself. What we just talked 7 about were the external adjustments to the model, if you 8 will. With regard to the working of the model itself, as 9 Dr. Peseau points out, the model is sensitive to the 10 selection of water years used and while you disagree with 11 Dr. Peseau about the appropriate years, I take it you 12 don't have any objection to that statement, do you? 13 A Which statement is that? I'm sorry. 14 Q That is that the model is very sensitive to 15 the selection of water years. 16 A The outcome, yes, it is sensitive to the 17 selection of water years. 18 Q Now, and you've already talked with 19 Mr. Shurtliff about your continued support of the 60-year 20 period. Let me deal with some of the specifics. On 21 page 3, lines 21 through 24 of your rebuttal testimony, 22 if you would go there. 23 A I'm there. 24 Q And here you quote Dr. Peseau as saying -- 25 well, first of all, you said starting at 18 and 19, 489 CSB REPORTING NORWOOD (X) Wilder, Idaho 83676 Avista 1 "Mr. Peseau has made a number of statements in his 2 testimony that are misleading," and then you go on to 3 give an example and you quote his statement about the 4 60-year water years producing the highest possible figure 5 for test year net power supply expenses; correct? 6 A I see that, yes. 7 Q And then you go on to say on the next page, 8 "This statement simply is not true." Now, I can't think 9 of any other way to ask this, Mr. Norwood, but isn't your 10 quotation of Dr. Peseau taken out of context? Doesn't he 11 in fact -- isn't it very clear that of the highest 12 possible years what he's referring to is even numbered, 13 decade-long years ending in '88? 14 A I can't speak to what he intended. What I 15 see that he wrote there was that he said that it happens 16 to produce the highest possible figure for power supply 17 expenses. 18 Q Yes, but it was of a selected number of 19 years that he was examining, that is, the 20-year through 20 60-year possibilities by decade ending in 1988; isn't 21 that true? 22 A Mr. Ward, all I can do is respond to the 23 statement which clearly states that he indicated that our 24 selection of years produced the highest possible figure 25 and the numbers show that it does not produce the highest 490 CSB REPORTING NORWOOD (X) Wilder, Idaho 83676 Avista 1 possible figure. 2 Q Well, let's go on to that. As you say, you 3 cite your page 1 of Exhibit 23, you say, "There are many 4 combinations of the same water record that would produce 5 a higher level of power supply expenses"; do you see this 6 testimony on lines 4 through 5 of page 4? 7 A Yes, I do. 8 Q And if I go over to Exhibit 23, what I find 9 there, first of all, on the first line on page 1 of 10 Exhibit 23 is your proposed water year; correct? 11 A That's correct. 12 Q And then I see six other water years that 13 you point out produce net purchases greater than the 14 proposed water year; is that correct? 15 A Yes. 16 Q All right. Now, are those the ones that 17 you -- are those all that produced greater net purchases? 18 A You could pick any combination. There's a 19 number of combinations that could be used and, again, it 20 depends on whether you want to limit it to five-year 21 increments, one-year increments. The point here is that 22 Mr. Peseau stated it would produce the highest possible 23 result. I simply here provided a few numbers to indicate 24 that it did not produce the highest possible figure. 25 Q Let me ask you this: It appears to me that 491 CSB REPORTING NORWOOD (X) Wilder, Idaho 83676 Avista 1 what you've done given the other water intervals that 2 you've cited, that is, '29 through '48, that stays on the 3 decade-long usage that we used before, but 1934 is a 4 different starting year, '29 -- I mean '28 is a different 5 starting year. It appears to me that you've looked at 6 all other potential water combinations of 20, 30, 40, 50 7 or 60 years in duration in that period; is that correct? 8 A Mr. Ward, I simply chose a number of years 9 that would have produced a higher set of power costs to 10 point out the fact that our proposal does not produce the 11 highest possible set of power costs. 12 Q I understand that, Mr. Norwood, but I'm 13 trying to ask how you identified those years. It looks 14 to me as if you ran every possible combination between 15 '28 and '88 in 20-, 30-, 40-, 50- and 60-year 16 increments; is that true? 17 A I'm sure I did not run every possible 18 combination. I ran a number of them, a few that are 19 listed here, just to demonstrate that the power costs 20 were not the highest that we filed. It was unnecessary 21 for me to run every possible combination. 22 Q Mr. Norwood, you knew intuitively which 23 years to go find that would show you a higher power 24 supply cost? 25 A It doesn't take long to run the numbers to 492 CSB REPORTING NORWOOD (X) Wilder, Idaho 83676 Avista 1 know what results it will produce. 2 Q Isn't it true it doesn't take long with the 3 model either to run all of the combinations of those 4 years? 5 A As I said, there wasn't a need to run every 6 possible combination. 7 Q So we don't know as we sit here -- well, 8 how did you know, for instance, to go look for 1934 9 through 1963 as a starting period? 10 A Mr. Ward, I simply had all the water record 11 data put on my machine in an Excel spreadsheet and just 12 started going through the numbers to find out which years 13 would produce a higher result and I found some that did 14 and I listed them here simply to point out that 15 Mr. Peseau's statement is not accurate. 16 Q All right; so you did run all the water 17 years? 18 A I listed the water years from 1928 to '88 19 there and then I simply started at some point and started 20 finding years that would produce higher results. 21 Q When I did the little scratching on a piece 22 of paper when I first looked at this exhibit, 23 Mr. Norwood, and tried to figure out how many 24 combinations one would get if you did exactly that, used 25 a spreadsheet with all of the combinations in these 493 CSB REPORTING NORWOOD (X) Wilder, Idaho 83676 Avista 1 years, I got 100 combinations. Does that sound right to 2 you? 3 A I did not run all possible combinations 4 because it was not necessary to run them. I had all the 5 water years listed, I identified a number that would 6 produce higher power costs and listed them here on the 7 exhibit. 8 Q You didn't necessarily, then, list all the 9 ones that provided lower power costs, did you? 10 MR. MEYER: I object. We have truly 11 covered this ground three times. Now, the witness is 12 very patient, more patient than I am, in responding to 13 repeated questions about what he did and didn't do. He 14 didn't purport to do a complete inventory of all the 15 possible combinations. He provided some illustrations as 16 he took a first pass at the modeling, but he didn't 17 purport to do all possible permutations. I object. I 18 think we've covered this ground. 19 COMMISSIONER SMITH: Mr. Ward. 20 MR. WARD: Madam Chair, I think probably 21 the point is obvious, but let me pursue it anyway. The 22 whole point of a normalization exercise is to draw in 23 fact conclusions about what is normal, so if you run a 24 spreadsheet and say, well, I can find six that are 25 higher, that, first of all, doesn't establish anything 494 CSB REPORTING NORWOOD (X) Wilder, Idaho 83676 Avista 1 unless we know how many were lower and I find it very 2 difficult to believe that there was not in fact a 3 compilation of the total data and that these years were 4 selected out. 5 MR. MEYER: The witness has testified under 6 oath there was not and that's where the record stands. 7 COMMISSIONER SMITH: Well, Mr. Meyer -- 8 MR. WARD: I'll withdraw the question and 9 ask it a different way. 10 Q BY MR. WARD: Mr. Norwood, you didn't look 11 at any other years other than these six; is that what 12 you're telling us? 13 A There may have been one or two other 14 periods that I chose to look at, but there wouldn't have 15 been more than one or two other periods that I chose to 16 look at. Again, the point was to show that what we filed 17 did not produce the highest power cost. I certainly did 18 not go through every iteration to find out every single 19 series of numbers that would produce higher or lower 20 power costs. 21 Q Mr. Norwood, let me ask you this question: 22 If you wanted to really refute what Dr. Peseau said, 23 wouldn't it be a simple matter to run all 100 24 combinations of the water year for this period and show 25 that roughly 50 were higher and 50 lower than your water 495 CSB REPORTING NORWOOD (X) Wilder, Idaho 83676 Avista 1 year, that would prove your point, wouldn't it? 2 A Mr. Ward, I guess I could have chosen a 3 number of methods to respond to Mr. Peseau. His 4 statement to me was very clear. His statement was that 5 we chose the highest set of power costs based on the 6 hydro data. This is the method that I chose. All I 7 would have to have listed here was one set of numbers 8 which would have shown that the data is inaccurate. I 9 chose to provide several as opposed to one. 10 Q Mr. Norwood, I can't help but being 11 intrigued by the idea of how you knew where to look for 12 some potentially higher years, so let me ask you this: 13 All of the six examples you give start with the years 14 1928 through 1934. What historical cataclysmic weather 15 event in the United States leaps to mind when I tell 16 you -- when I cite the years 1928 through 1934? 17 A I've worked with these numbers for many 18 years, Mr. Ward, so I'm familiar with the years that are 19 above normal, whatever normal is, streamflow conditions 20 and those that are below normal, so it doesn't take a 21 rocket scientist to figure out which years would produce 22 high results, which ones that would produce lower 23 results. 24 Q I take it from your answer to my question 25 you've never read the Grapes of Wrath? 496 CSB REPORTING NORWOOD (X) Wilder, Idaho 83676 Avista 1 A If I have, it's been many years. 2 Q All right, let's go on. Go back to page 6 3 of your testimony, your rebuttal testimony, and I'll be 4 on your rebuttal testimony from now on, Mr. Norwood. 5 A I'm there. 6 Q Here you're talking about, to shorten it 7 just a little bit, Dr. Peseau mentioned an Idaho Power 8 case in which the use of water years was hotly disputed 9 and you've pointed out in rebuttal that in the '92 and 10 '94 case, the use of a longer water year, that is, 65 11 years, was not a contentious issue in Idaho Power's 12 cases. Do you recall that testimony in substance? 13 A Yes, I do. 14 Q Do you know why it was not a contentious 15 issue? 16 A No, I do not. 17 Q Did you look at the transcript of either of 18 those proceedings or, for that matter, of the 265 19 proceeding Dr. Peseau cites? 20 A No, I reviewed the orders in the case. 21 Q Okay. Go to page 8. Now, on lines 17 22 through 20, you say there that the Company's filing 23 provides to retail customers the full benefits of all 24 secondary purchase and sales and you go on to the end of 25 that sentence. Do you see that? 497 CSB REPORTING NORWOOD (X) Wilder, Idaho 83676 Avista 1 A Yes, I do. 2 Q Does it provide the full benefit of all 3 actual secondary purchases and sales for system load or 4 otherwise or does it provide a normalized cost of supply? 5 A The general rates are based on a normalized 6 set of numbers. Through the power cost adjustment that 7 we have in place, the actual prices of the market 8 conditions as well as the hydro conditions are tracked. 9 Q Now, maybe I didn't make my question 10 clear. In a test year or, for that matter, in any period 11 of time, there will be actual purchases and sales made 12 for system purposes; correct? 13 A Yes. 14 Q And in fact, do any of those numbers other 15 than the contract adjustments that we've talked about, on 16 a short-term basis, do any of those actual transactions 17 have any impact whatsoever on your power supply 18 adjustment in this case? 19 A I guess I'm going to need you to restate 20 the question. I didn't follow exactly what you're after. 21 Q I'm still being too clumsy. Isn't it true 22 that the actual transactions, the actual -- let's use 23 short term because it's not as complicated -- the actual 24 short-term transactions for system benefit in this 25 proceeding, the actual transactions that occurred during 498 CSB REPORTING NORWOOD (X) Wilder, Idaho 83676 Avista 1 the year are irrelevant, the model supplies the answer; 2 isn't that the case? 3 A It's true the '97 actual numbers in this 4 case have been adjusted to reflect normal conditions for 5 this rate case, that's true. 6 Q And that's one of the reasons we get to the 7 kind of results that are reflected in your Exhibit No. 6 8 where the actual short-term purchased power is 9 $191 million and change, if you look at line 1 of page 10 1 -- 11 A I see that. 12 Q -- and of that $191 million and change, 13 $174 million is adjusted out. 14 A That's correct. 15 Q And it's adjusted out because the model 16 predicts under these conditions only the 16,268,000 that 17 results. 18 A That's correct. 19 Q Did the 174 million that's adjusted out 20 flow through the PCA? 21 A No, it did not. Let me back up. Part of 22 it would have. To the extent that the 174 million 23 includes changes in costs related to hydroelectric 24 generation, it would. There are some changes related to 25 thermal costs that are -- that flow through the PCA and 499 CSB REPORTING NORWOOD (X) Wilder, Idaho 83676 Avista 1 market pricing would also be picked up, so some of it 2 would be flowed through the PCA. 3 Q Those would be relatively minor, maybe 4 relatively quite minor would be a better description, of 5 the total of 174 million; isn't that true? 6 A In relation to the 174 million, yes. 7 Q Okay, and the same thing happens, just so 8 there's no confusion, on power sales; isn't that true? 9 We have the same very sizeable adjustments. If you look 10 at page 2 of 4, line 85 of Exhibit 6, actual was 11 192,357,000 out of which 182 million, all but 10 million, 12 is adjusted out? 13 A That's correct. It's important to note, 14 too, that if you look at the two adjustments, you have 15 175 million reduction in purchases and 182 million or 16 183 million reduction in sales on a net basis. You're 17 looking at, what, if I do the math right, about $8 18 million on a net basis. 19 Q All right, and so to return to your quote 20 on page 8, lines 17 through 20, where you say the retail 21 customers get the full benefits of all secondary purchase 22 and sales transactions, with respect to the actual 23 secondary and sales transactions, wouldn't it be correct 24 to say that they do not get the benefit of the actuals in 25 this rate case and they did not get the benefit in the 500 CSB REPORTING NORWOOD (X) Wilder, Idaho 83676 Avista 1 PCA adjustment? 2 A On page 8, I mentioned that they get the 3 full benefit of all purchase and sales transactions 4 related to the Company's system, the generating 5 resources, as well as any purchases made related to 6 serving load. 7 Q But it's only the model benefit; isn't that 8 true? 9 A It is the model benefit which we use to 10 reflect the weather corrected loads, the normalized hydro 11 generation, all the contract changes and that's the 12 purpose for the dispatch model is to reflect all those 13 changes in the power costs that we include for ratemaking 14 purposes. 15 Q And in fact, as we go forward, assuming the 16 PCA stays in place, it will always be true that actual 17 sales, actual prices will not matter, that what we will 18 get is the model's normalized PCA adjustment? 19 A No, on a going forward basis for the PCA 20 the actual prices are used to determine the price that 21 the surplus would be sold at or the price that purchases 22 were made to serve load, so there's a price component 23 that is built into the PCA based on actual prices. 24 Q But the actual sales are not? 25 A The actual sales prices are also included 501 CSB REPORTING NORWOOD (X) Wilder, Idaho 83676 Avista 1 in the PCA. The speculative trading that I've talked 2 about in my rebuttal and in my direct testimony is not 3 included. 4 Q Now, I want to go on to that speculative 5 trading. With respect to -- basically, you go on here 6 shortly to rebut what I'll characterize as Dr. Peseau's 7 concern about the commingling of all sales and purchases, 8 speculative, system, whatever, and starting on page 9, 9 you're responding to that argument, are you not? 10 A Yes. 11 Q And starting on page 9 at lines 15 through 12 16, you're asked, "How are the commercial secondary 13 transactions different than the system secondary 14 transactions?" And starting at lines 15 and 16 you say 15 they have these characteristics and you call off five 16 characteristics, do you not? 17 A Yes. 18 Q Of these characteristics -- well, let me 19 ask it this way: Of these characteristics, what you're 20 really describing are the situations that should 21 determine whether we regard a purchase or sale as 22 speculative or system basis; isn't that correct? Strike 23 that, that's a bad question. Let me ask it another way. 24 Isn't it fair to say that the way we should 25 look at your conditions or your indicia of speculative 502 CSB REPORTING NORWOOD (X) Wilder, Idaho 83676 Avista 1 transactions goes something like this: If you'd turn 2 over to page 10, line 4, you'll see the fifth 3 characteristic and you say, "Shareholder capital is 4 placed at risk...", wouldn't it be fair for me to 5 rephrase that and say shareholder capital should be 6 placed at risk if the Company is to take any credit for a 7 transaction? Do you have any quarrel with that? 8 A I guess I'm not sure what you're saying 9 there. For these transactions the shareholder dollars 10 are at risk. If there's a loss on the sale, the 11 shareholders absorb it. If there's a gain on the sale, 12 they retain it. 13 Q Let me try it another way, Mr. Norwood. 14 Let's go to your rebuttal Exhibit 23, page 11 of 13. 15 A I'm there. 16 Q Here you've introduced into evidence two 17 deal tickets and they're dated for consecutive days. The 18 first one was June 4, '99, the second one June 5; is that 19 correct? 20 A May. 21 Q I'm sorry, May 4th and 5th, and with these 22 deal tickets, what they reflect is that the Company sold 23 short on the 4th 25 megawatts firm, on peak power and on 24 the next day it covered its short position with a 25 purchase of 25 megawatts of firm, on peak power. Do you 503 CSB REPORTING NORWOOD (X) Wilder, Idaho 83676 Avista 1 see that? 2 A Yes, I do. 3 Q Now, let me ask you this: Is there any -- 4 first of all, by the way, both of these deal tickets have 5 Avista Utilities at the head, can I assume that they were 6 done by the generating and resources folks? 7 A That is correct. 8 Q So these are on the regulated side or 9 within the regulated entity; is that correct? 10 A That's correct. 11 Q So they're within the regulated entity, do 12 you see anything on the face of the ticket on the 4th 13 that indicates that this is a speculative transaction or 14 otherwise belongs to the nonregulated sector? 15 A No, there's nothing indicated. 16 Q Anything on the face of the ticket that 17 indicates that only shareholder capital is at risk here? 18 A No, nothing indicated. 19 Q Is there anything that indicates that it's 20 unrelated to transactions used to serve retail load? 21 A Nothing indicated here. 22 Q And I could walk through the rest of them, 23 but let me -- 24 A I think it's important to note on one of 25 the others that you're skipping over and that is the 504 CSB REPORTING NORWOOD (X) Wilder, Idaho 83676 Avista 1 reliance on system resources. For each one of these deal 2 tickets, there's a confirmation that's put together and 3 the vast majority of the deals that we do are done under 4 our WSPP agreement, Western Systems Power Pool agreement, 5 which we're a party to. In that agreement, it identifies 6 the type of product here, it's indicated firm. Under 7 schedule C of that agreement, there's two separate 8 categories which identify the products that are being 9 traded under that WSPP agreement. 10 One is a financial firm product which is 11 not backed by firm resources. It has a liquidated 12 damages provision to it so that you're not obligated to 13 deliver, but you are obligated on a financial basis to 14 make whole the other party if we don't deliver. The 15 other product is a system firm resource which is required 16 to be backed by reserves. These products that we're 17 trading here are a financial firm product and they are 18 not backed by the Company's resources. 19 Q But, of course, only you make that 20 declaration to the WSCC; isn't that true? 21 A At the time the deal is done the parties 22 agree as to what the terms and conditions are. 23 Q Now, I want to follow up with your thought 24 that no system resources are used, and maybe the best way 25 to do this is if you just briefly turn to the next page 505 CSB REPORTING NORWOOD (X) Wilder, Idaho 83676 Avista 1 of the exhibit, there you have the organizational chart 2 for resource optimization. Do you see that? Isn't that 3 generation and resources as it was otherwise called in 4 the testimony? 5 A I believe I referred to it as resource 6 optimization, if I'm not mistaken, in my testimony. 7 Q Yes, I understand, but I'm trying to relate 8 it to the overall organizational scheme of the Company. 9 A It is within the regulated utility if 10 that's your question. 11 Q In Mr. Matthews' corporate organizational 12 structure on page 5, he shows the corporate, a corporate 13 entity or a corporate regulated line as generation and 14 resources line of business. Is resource optimization 15 within that line of business? 16 A Yes, it is. 17 Q But I take it, it's not all of that line of 18 business? 19 A That's correct. 20 Q So it's a subset of generation and 21 resources? 22 A Yes. 23 Q And I take it this is where the -- this is 24 the organization that makes the deals, in short, on power 25 supply? 506 CSB REPORTING NORWOOD (X) Wilder, Idaho 83676 Avista 1 A Yes. 2 Q All right. Now, let me ask you a 3 hypothetical paranoia question. What I see on this first 4 deal ticket which sold short for 25 megawatts, let us 5 suppose that I'm the trader who sold that short sale and 6 I decide not to cover it for some considerable period of 7 time, I think the market is going to go down, that 8 happens sometimes, doesn't it? 9 A Yes, it does. 10 Q And I decide that I'm going to sit on that 11 for 30 days and then cover my short. Meantime, let's say 12 we're in a shoulder month and Washington Water Power for 13 its own system purposes is holding water in preparation 14 for its anticipated winter peak whenever possible. Do 15 you have that in mind? 16 A Okay. 17 Q Now, I made this short sale and maybe it's 18 not 25 megawatts, plug in any number you want, and all of 19 a sudden, during the 30 days I'm trying to sit on it 20 suddenly the market goes berserk and I have an extreme 21 exposure, pick a number, millions of dollars on this 22 sale. There's two possibilities it seems to me. One, of 23 course, I can make the necessary purchase and cover and 24 lose millions of dollars. The other possibility, of 25 course, is that Avista could cover with the use of some 507 CSB REPORTING NORWOOD (X) Wilder, Idaho 83676 Avista 1 of that water resources that's not necessarily required 2 immediately otherwise. Don't you think that situation is 3 a powerful temptation to covering with utility resources? 4 A I guess it's a long question. I'm thinking 5 of how to respond to all those issues that are built in 6 there. First of all, our scheduler will run the system 7 and optimize the system, optimize the use of hydro 8 generation to get the greatest value for that as well as 9 optimize the thermal generation. Once -- as far as the 10 possibility of costs getting out of control, we have risk 11 parameters that are pretty tight within that resource 12 optimization group which does not allow them to get into 13 the position to where you can lose many millions of 14 dollars. 15 MR. MEYER: Do you have the thrust of his 16 question in mind? 17 THE WITNESS: Maybe I don't, exactly. 18 MR. MEYER: Would you rephrase it, please? 19 MR. WARD: Well, let me go on. 20 Q BY MR. WARD: Let me ask you where 21 physically the resource optimization folks are located 22 with regard to the dispatch center. Are they on the same 23 floor? 24 A The dispatch, the power system dispatch, 25 group is within the resource optimization group. 508 CSB REPORTING NORWOOD (X) Wilder, Idaho 83676 Avista 1 Q Okay; so I take it in order to talk to the 2 power dispatchers if I'm a marketer within that group, I 3 can walk down the hall and have a discussion with them? 4 A The pre-scheduled group as well as the 5 real-time hour-to-hour people are in the same room. 6 Q Let me give you another hypothetical. 7 Let's suppose that I being the sharp trader I am in 8 resources see an opportunity to make a very attractive 9 purchase of resources at 10 mills. That can happen; 10 correct? 11 A Yes. 12 Q We hope it happens a lot. I make that 10 13 mill purchase and now I find myself in the situation a 14 few days later, a situation of my dreams, exactly of my 15 dreams, which is that the prices triple and I can sell it 16 for 30. Meanwhile, though, the price for -- the system 17 is itself in deficit. Now, there's many permutations of 18 that possibility of combinations, but isn't it -- and if 19 the system is in deficit and power is dear, let us say 20 that on the original purchase I purchased for the system, 21 but now I can sell it for 30 mills and the system which 22 is short, of course, may have to go buy if I do that, 23 where are the cops on the corner that would prevent that 24 from happening in the resource optimization group? 25 A The way the ratemaking is set up now, I 509 CSB REPORTING NORWOOD (X) Wilder, Idaho 83676 Avista 1 guess the cops are automatically in place. Let me 2 explain that. For normalized ratemaking, we provide 3 customers the full benefits of the operation of the 4 resources, hydro, thermal, any purchases that serve load 5 as well as sales and surplus. When you look at the power 6 cost adjustment, the power cost adjustment includes any 7 changes from the rate case amounts, the normalized 8 amounts, related to hydroelectric generation changes as 9 well as market pricing. 10 To the extent that the Company chooses to 11 enter into additional purchases or additional sales that 12 are unrelated to serving load, they will automatically 13 fall upon the shareholder. They will automatically fall 14 upon the shareholder whether it's a purchase or a sale if 15 the Company chooses to engage in additional transactions, 16 and by the way that rates are set, by the way the PCA 17 works, it automatically has a cop built in so that 18 shareholders are at risk. 19 Q I understand, but that's a function of the 20 model; correct? 21 A It's a function of the model and the 22 existing PCA and existing ratemaking which places the 23 shareholder at risk for any additional transactions that 24 it chooses to engage in. 25 Q All right, Mr. Norwood, let's go on a 510 CSB REPORTING NORWOOD (X) Wilder, Idaho 83676 Avista 1 little bit. 2 MR. WARD: May I have just a second? 3 COMMISSIONER SMITH: Certainly. 4 (Pause in proceedings.) 5 Q BY MR. WARD: Mr. Norwood, looking at those 6 deal tickets, let me ask you simply on the counter party 7 where I see Statoil -- 8 A Statoil. 9 Q Statoil? -- and ECI, can Avista's name 10 ever appear on those deal tickets in that blank? 11 A I do not believe we have transactions with 12 Avista on the electric side, Avista Energy on the 13 electric side, so no. 14 Q Avista Energy, did you understand me to 15 mean Avista Energy? 16 A Yes, and my response is we do not engage in 17 transactions with Avista Energy is my understanding. 18 Q Could Avista Utilities itself be listed 19 there? 20 A No, it would be buying or selling with 21 ourself, so no. 22 Q Okay. Now, on page 18 of your rebuttal 23 testimony, here you begin by -- at this point you're 24 responding to Dr. Peseau's proposal to allocate costs to 25 the commercial trading activities and let me just make 511 CSB REPORTING NORWOOD (X) Wilder, Idaho 83676 Avista 1 one thing clear, I trust it's been clear throughout, but 2 throughout this we're just talking about the commercial 3 trading activities within the regulated entity, are we 4 not? 5 A That's correct. 6 Q In other words, we haven't even proposed or 7 Dr. Peseau didn't even propose any allocation to the 8 unregulated side? 9 A That's my understanding. 10 Q Okay. Now, you take exception to that 11 proposal -- well, first of all, let me ask you, in your 12 original testimony, you didn't do any allocation to the 13 unregulated or the shareholder side at all of any 14 employee costs or any overhead costs, did you? 15 A That's correct. 16 Q And now in response you've come back with a 17 recommendation that basically allocates on a 18 jurisdictional basis $157,000 and change to the 19 shareholder trading activities -- 20 A No. 21 Q -- correct? 22 A No. I've identified the costs that could 23 be allocated to those activities, well, the costs that 24 could be assigned to the activity, but I also identified 25 in here some offsetting margins that a good case could be 512 CSB REPORTING NORWOOD (X) Wilder, Idaho 83676 Avista 1 made that should accrue to shareholders, so there are 2 some offsets that I've included here, but if a decision 3 were made to allocate A&G costs to commercial trading 4 activities, it would be in the neighborhood of these 5 costs that I've identified here. 6 MR. WARD: May I approach the witness? 7 COMMISSIONER SMITH: Certainly. 8 (Mr. Ward approached the witness.) 9 MR. WARD: I believe our next is 210. 10 (Potlatch Corporation Exhibit No. 210 11 was marked for identification.) 12 Q BY MR. WARD: All right, Mr. Norwood, 13 showing you what's been marked as Exhibit 210 for 14 identification, this is a supplemental discovery response 15 that I received, I guess, last Friday, and in this, if 16 you turn to the second page, you're asked a variety of 17 questions about actual labor and other expenses for 18 resource optimization. Do you see that? 19 A Yes. 20 Q And then I take it at the bottom 21 handwritten on the following pages is 1, 2, 3, 4, that's 22 not mine, that came from the Company as far as I know. 23 A That's correct. 24 Q What you've given us is resource 25 optimization labor totals for '97; everything other than 513 CSB REPORTING NORWOOD (X) Wilder, Idaho 83676 Avista 1 labor for '97 is No. 2; labor for '98 is No. 3; and other 2 than labor for '98 is on No. 4. Do you recall that? 3 A Yes, I do. 4 Q Now, it appears to me that if I add up '98, 5 for instance, that I get something over $9 million for 6 resource optimization costs in total; is that correct? 7 A Yes, that's correct. 8 Q And looking at your organizational chart 9 which is Exhibit 23, page 12 of 13, if I counted 10 correctly, I counted 30 positions in that chart; is that 11 roughly correct? 12 A I haven't counted them. One thing that I 13 should point out is that this organization chart is just 14 for the electric group for the resource optimization. In 15 fact, I believe it includes all of them within the 16 electric organization. There's a gas group that's also 17 included in resource optimization which isn't listed here 18 and their costs are included in the resource optimization 19 in this Exhibit 210. 20 Q How many bodies, roughly, positions, are in 21 the gas group? 22 A We have seven or eight individuals in the 23 gas group. 24 Q Okay; so if I'm right that there's 30 in 25 the electric, give or take, seven or eight, an additional 514 CSB REPORTING NORWOOD (X) Wilder, Idaho 83676 Avista 1 seven or eight, are gas; is that correct? 2 A Yes. 3 Q All right. Now, these 9 million in total 4 costs -- well, first of all, when you on rebuttal make 5 your estimate of the appropriate allocation of costs to 6 the speculative trading activity, you say there is 7 only -- we would need all but four people in this 8 department anyway even if we didn't have speculative 9 trading; correct? 10 A Yes. 11 Q My first question regarding -- well, and 12 then you go on to use that, so to speak, those four 13 people to determine their incremental labor cost and 14 incremental overhead and that is the basis for your 15 adjustment; correct? 16 A In general terms, yes. 17 Q So, for instance, in Exhibit 23, page 13, 18 here's where you've made the calculations and rather than 19 allocate a pro rata share of the total office space, you 20 just assume that the incremental needs would be 90 square 21 feet for employee and then you take that pro rata share 22 of the total for an employee; correct? 23 A Yes, the office space was figured based on 24 an equivalent of four positions. 25 Q All right. Now, if you would need 26 of 515 CSB REPORTING NORWOOD (X) Wilder, Idaho 83676 Avista 1 these 30 positions to operate the utility even without 2 speculative sales, how is it that this department was not 3 organized until 1996? 4 A The department was reorganized in or, 5 actually, I guess the name was adopted or the group was 6 assembled in 1996. The functions that exist today 7 existed back then. Maybe we could walk through those if 8 you'd like. The real-time operators, we have five of 9 those. Someone has to be there 24 hours per day. We had 10 those before resource optimization existed, we have those 11 today. We have a pre-scheduler which is needed. We have 12 people that do the padding on transmission. We have 13 accountants. 14 For the most part, all of these positions, 15 and as I mentioned in my rebuttal testimony, there's 16 really only one individual that's there that's dedicated 17 solely to "trading," the remainder of them are there 18 primarily dealing with transactions that relate to 19 serving core load and operating the system, but we 20 recognize that some of their time is dedicated to or is 21 used to support the commercial trading activities, so 22 we've recognized that and allocated some of their time to 23 that activity. 24 Q So of the short-term purchases of -- let me 25 avoid going back and looking at the number, but roughly 516 CSB REPORTING NORWOOD (X) Wilder, Idaho 83676 Avista 1 $182 million and $170 million, roughly, is excluded 2 because it's speculative activity, but you would only 3 need four of these 30 people to do that 170 million if 4 90 plus percent got excluded? 5 A That's correct. 6 Q Is that what you're telling me? 7 A That's correct. 8 Q All right. Now, even if you -- 9 A And that's illustrated, like I mentioned, 10 in my rebuttal testimony in how that through a single 11 transaction you can have a significant volume of both 12 dollars and megawatt-hours just from a single transaction 13 and that's what contributes toward the large numbers, the 14 large adjustments that were made, but, again, you have to 15 look at it on a net basis. If you have a sale and a 16 purchase, you're going to have a large purchase and a 17 large sale. On a net basis, though, the adjustment to 18 short-term sales and short-term purchases is only 19 $8 million. 20 Q I understand that, Mr. Norwood. Returning 21 to Exhibit 210, since we have about $9 million in total 22 '98 costs, even if I accept your contention that the 23 four out of 30 positions is correct, if it was -- if 24 those costs were fully allocated, they would surely be 25 considerably more than the amounts you've allocated on an 517 CSB REPORTING NORWOOD (X) Wilder, Idaho 83676 Avista 1 incremental basis; isn't that true? 2 A I don't know whether that would be true or 3 not. I would have to run the numbers. 4 Q Well, you can't look at it and see whether 5 that would be true? 6 MR. MEYER: Excuse me, by "fully 7 allocated," do you mean all of the time of the four 8 individuals or the appropriate share of the time of each 9 of the four individuals? 10 MR. WARD: I mean if the total costs of 11 this department were allocated, if it's to be allocated 12 on the basis of individuals, if all were allocated to all 13 individuals equally, I think the witness understood what 14 I was asking. 15 THE WITNESS: One thing you have to 16 consider is that this Exhibit 210 includes costs 17 associated with the gas business, too, and you'd need to 18 go through and see what the specific costs are that are 19 directly related to the natural gas business and then 20 identify what's related to electric and that would 21 require a significant amount of time to do that. 22 What we've done is taken a look at, as I 23 showed on pages 12 and 13 is to take a look at, 24 specifically what functions, what positions support the 25 commercial trading activities and tried to isolate those 518 CSB REPORTING NORWOOD (X) Wilder, Idaho 83676 Avista 1 dollars that are directly related to those commercial 2 trading activities. This other approach is kind of a 3 shotgun approach and it requires a lot more analysis than 4 just taking a ratio of employees to dollars to figure out 5 whether the number fits or not. 6 Q BY MR. WARD: Well, Mr. Norwood, you're 7 making this more complicated than is necessary. All I'm 8 asking you is you admitted that you allocated the cost to 9 the competitive side on an incremental basis. All I'm 10 asking you is if we didn't allocate it on an incremental 11 basis, if it was allocated on a pro rata basis based on 12 employees, even adding in seven or eight employees for 13 gas, with roughly $9 million at issue, you're going to 14 get a much larger number than your $471,000 on a system 15 basis, are you not? 16 A I don't know that that would be true. 17 Q I'm only asking you for a mathematical 18 calculation. Divide 9 million by 38 and then times 19 four. 20 A You don't know how much of those costs 21 should be dedicated and allocated to the gas group, so 22 it's not as simple as a simple calculation to allocate 23 between employees. It may be that the costs associated 24 with running the gas group are higher than the electric 25 or vice versa and that could have a material impact on 519 CSB REPORTING NORWOOD (X) Wilder, Idaho 83676 Avista 1 the result and that would be a very, very simplified 2 approach. 3 What we've chosen to do is actually look at 4 the people that are dedicated to or have some 5 relationship to the commercial trading activities and 6 identify those costs specifically and I think the number 7 that we've come up with, which is about a half million 8 dollars, is a good representation of what it takes to 9 support those activities. 10 Q Let me ask you this: With that incremental 11 analysis, aren't you in fact supporting some of what 12 Dr. Peseau has expressed concerns about and that is how 13 the Company will choose to allocate revenues, expenses, 14 et cetera, when there's a commingling of shareholder and 15 ratepayer interests? 16 A I guess I don't see that there's a problem 17 there. Like I said before, with the transactions 18 themselves, the cops are already in place which 19 automatically put the Company at risk for any of these 20 transactions they choose to enter into. For the 21 allocation of costs or identification of costs that are 22 related to these activities, what we really should do is 23 exactly what we did in this analysis here and that is 24 identify the costs that support this activity and 25 allocate them to that activity. 520 CSB REPORTING NORWOOD (X) Wilder, Idaho 83676 Avista 1 You mentioned the incremental and I'm not 2 sure what you mean by that, but we didn't look 3 specifically at incremental costs. We looked at 4 positions which support those activities and allocated 5 the time of those individuals, so if that's what you mean 6 by incremental, then that's what we did. I don't want to 7 get lost in the terms. 8 Q Well, I don't want to pursue that. We'll 9 just go back down the same road. Let me ask you this, 10 Mr. Norwood: Quite apart from the interests of the 11 ratepayers, there are competitive concerns that the 12 Commission should take account of in this whole 13 commercial trading world, are there not; in other words, 14 if we're going to have a competitive market, then it 15 should be a fair competitive market; would you agree? 16 A In general terms, yes. 17 Q Suppose Enron or Idaho Power showed up on 18 your doorstep tomorrow and said you've still got some 19 space there and I'd like to put four more employees in 20 there paying only for 90 square feet of space, what's 21 your response going to be? 22 A I think that goes beyond what we're 23 proposing here in this case and the operations that are 24 within the resource optimization. That's more of a 25 policy decision on where we go to the future with these 521 CSB REPORTING NORWOOD (X) Wilder, Idaho 83676 Avista 1 types of operations. The operations that we have within 2 resource optimization related to commercial trading 3 activities is really a very limited operation when you 4 compare it to other full-scale trading operations. The 5 Company has chosen to put some pretty restrictive limits 6 on the open positions which limits the exposure to the 7 Company. 8 Q And I take it -- you got a little feel 9 there, I think, Mr. Norwood, but I take it your response 10 to Idaho Power and Enron would be the same if they said 11 and oh, by the way, in addition to the 90 square feet, we 12 want to use your computers, your copying facilities, your 13 telephones, et cetera for free? 14 A We have costs included in here for copiers 15 and all of those things. I mean, this is similar to 16 other analyses when you identify costs associated with 17 certain activities. 18 Q Would you have any objection if they took 19 advantage of those very similar analyses? 20 A I think the question is a broader question 21 than that and goes beyond whether we would let them use 22 our floor space and that sort of thing. If you look at 23 the analysis, I think it's very clear, we've identified 24 the costs associated with those activities, including 25 floor space and the furniture and copiers and telephones 522 CSB REPORTING NORWOOD (X) Wilder, Idaho 83676 Avista 1 and all those things. 2 MR. WARD: That's all I have. Thank you. 3 COMMISSIONER SMITH: Thank you, Mr. Ward. 4 MR. MEYER: Would it -- 5 COMMISSIONER SMITH: Mr. Meyer? 6 MR. MEYER: In deference to the witness, 7 he's been on the stand for quite some time, could we have 8 just a short recess? 9 COMMISSIONER SMITH: Mr. Norwood, would you 10 like a five-minute break? 11 THE WITNESS: Yes. 12 COMMISSIONER SMITH: All right. 13 THE WITNESS: Thank you. 14 (Recess.) 15 COMMISSIONER SMITH: I think we're ready 16 for Mr. Woodbury's cross of Mr. Norwood. 17 MR. WOODBURY: Thank you, Madam Chair. 18 19 CROSS-EXAMINATION 20 21 BY MR. WOODBURY: 22 Q Mr. Norwood, what Mr. Ward identified as 23 Exhibit 210 was a follow-up to a Staff production 24 request, No. 61, and that was the annual operating 25 expenses incurred to undertake the short-term purchase 523 CSB REPORTING NORWOOD (X) Wilder, Idaho 83676 Avista 1 and sales activities; is that what those pages were? 2 A The pages that were included include the 3 operating costs for the resource optimization department, 4 which, as I stated before, includes all the resource 5 optimization, electric and gas. Within that would be the 6 costs associated with short-term transactions. 7 Q Does the Company have any information 8 regarding the cost of conducting secondary system 9 transactions before 1997? 10 A Not to my knowledge, no. 11 Q Then there is no precise way to tell other 12 than the estimate presented in rebuttal how expenses 13 associated with secondary purchases and sales have grown 14 with respect to the growth of secondary transactions? 15 A Yeah, I think that's true. It would 16 require a lot of analysis to go back and try to identify 17 costs, in part, because the group has changed over time. 18 Some groups have been brought together within resource 19 optimization that were previously in different parts of 20 the Company. 21 Q In response to Staff production request 22 61(d) or Exhibit 210, you show an expense increase for 23 the resource optimization group of about 11 percent 24 between '97 and '98. Why would an increase of that 25 magnitude not be the result of increased speculative 524 CSB REPORTING NORWOOD (X) Wilder, Idaho 83676 Avista 1 transactions? 2 A There are a number of changes that have 3 occurred from '97 to '98 and it would require a detailed 4 analysis to see what the changes in costs were. In fact, 5 in the middle of '97, that's the time Avista Energy was 6 formed, so, therefore, there were some changes in 7 personnel at that time. Again, you'd have to go back and 8 identify and do some detailed analysis on these costs to 9 see which costs changed. 10 Q The resource optimization group before it 11 was formed and then named, were they actually performing 12 those services back in 1996? 13 A Which services are you referring to? 14 Q Speculative commercial transactions. 15 A I wasn't directly involved in the group at 16 that time. There may have been some. I don't know how 17 much. 18 Q You show in response to Staff production 19 request No. 61(a) the actual annual short-term purchases 20 and sales for 1993 to 1998. Do you recall that response? 21 A Yes, I do. 22 Q And essentially, you show that for 23 short-term purchases in dollars per millions, in '93 it 24 was 31.8; in '94, 38.3; '95, 29.2; and then in 1996 it 25 shoots all the way up to 99.8; in 1997 it shoots up to 525 CSB REPORTING NORWOOD (X) Wilder, Idaho 83676 Avista 1 191.1; and in 1998, 347.1. 2 A Yes, I see those numbers. 3 Q Is it reasonable to look at the increases 4 that occurred in '96, '97 and '98 as related to the 5 speculative transactions? 6 A Yes, I'm sure that the bulk of those, the 7 vast majority of those, transactions and dollars would be 8 associated with the commercial trading activity. 9 Q Staff production request No. 6 asked the 10 Company's position regarding elimination of all 11 speculative trading of energy from the regulated Company 12 and I spoke about that earlier, I think, with 13 Mr. Matthews. In response, you state that the Company is 14 considering a proposal to consolidate the management of 15 its electric resources under Avista Energy, so the 16 question was with respect to speculative trading, your 17 answer was with respect to management of electric 18 resources, do you include speculative trading as electric 19 resources of the Company? 20 A No, the reference here to management of 21 resources would refer to the assets of the utility. The 22 speculative transactions would not fall into that 23 category. 24 Q Okay. Is it possible to move only the 25 speculative commercial trading activities from the 526 CSB REPORTING NORWOOD (X) Wilder, Idaho 83676 Avista 1 regulated Company to Avista Energy? 2 A I guess it really isn't a matter of moving, 3 it's a matter of choice as to whether you choose to do it 4 or not. Avista Energy does engage in speculative trading 5 or commercial trading activity, and as I've stated 6 earlier, Avista Energy does a lot more than what occurs 7 within the regulated utility. At this point in time the 8 Company does choose to engage at some level, but I also 9 mentioned that there are some pretty strict guidelines to 10 where the exposure to the Company is limited because of 11 the parameters that are set within the utility itself, so 12 if it chose to eliminate it at the utility, it wouldn't 13 be a matter of moving it, per se. It would be a matter 14 of choosing not to do it within the utility. 15 Q In your opinion, is there or can there be 16 any relationship between a speculative commercial 17 transaction of the Company in either meeting retail load 18 or selling energy from generating resources of the 19 Company? 20 A No. These transactions are unrelated to 21 the Company's existing resources and they're not related 22 to the purchases that are made to serve retail load. 23 Q There was some discussion earlier this 24 morning regarding the Company's sale of its interests in 25 Centralia. 527 CSB REPORTING NORWOOD (X) Wilder, Idaho 83676 Avista 1 A Yes. 2 Q And I think in Exhibit 9 showing your 3 dispatch model results you show the thermal fuel expenses 4 of $35 million; is that correct? 5 A Yes. 6 Q And of that, 22 million is related to 7 Centralia? 8 A Yes. 9 Q If Centralia were to be sold, the resultant 10 effect on short-term sales and short-term purchases would 11 be what? 12 A If Centralia were to be sold, then there 13 are a number of things that would occur and could occur. 14 You obviously would have a reduction in the fuel cost 15 that you see here. There would be changes to O&M cost, 16 changes to capital cost, including depreciation and 17 returns. There would be changes in transmission costs 18 associated with getting the power to our system. There 19 would also be changes in costs associated with whatever 20 power supply would be used to replace the Centralia 21 output, so there's a number of costs that would change if 22 a sale were to go through. 23 Q Would you expect to see a decrease in 24 short-term sales and perhaps an increase in short-term 25 purchases? 528 CSB REPORTING NORWOOD (X) Wilder, Idaho 83676 Avista 1 A Again, it would depend on the power that's 2 acquired to replace Centralia. One possibility is for us 3 to purchase the output of Centralia for some period of 4 time as part of the deal. We could buy some long-term 5 power as one option to replace the power, so it would 6 depend on the replacement power as to whether there would 7 be an increase or decrease in short-term transactions. 8 Q Have you run, made a run of your dispatch 9 model factoring Centralia out? 10 A I have not, no. 11 Q On page 18 of your direct testimony you 12 state, "Under normal streamflow conditions, the Company 13 is near load/resource balance." What do you mean by that 14 statement? 15 A If you take a look at -- if you basically 16 just stack up our retail loads and all of our wholesale 17 contract obligations and compare those, the rights we 18 have, the contracts delivered to us, and then you add in 19 generation from our hydro projects under what we use as 20 normal streamflow conditions and then you run your 21 thermal plants as we have in the dispatch model, then in 22 this case, we would end up approximately a net purchaser 23 of 50 average megawatts, but as has been pointed out 24 already, some of those plants have been displaced; that 25 is, it was cheaper to buy non-firm energy than to run the 529 CSB REPORTING NORWOOD (X) Wilder, Idaho 83676 Avista 1 thermal plants based on the dispatch model, but if you 2 add back in the availability from the thermal plants, 3 then you end up with the Company roughly having firm 4 resources equal to firm loads. 5 Q And you're factoring Centralia into those 6 resources? 7 A Yes, I am. 8 Q There was a Rob Strenge, is he a Company 9 employee? 10 A Yes, he is. 11 Q He indicated that Centralia serves about 12 12 percent of the utility's power needs; is that an accurate 13 statement? 14 A There's approximately 200 megawatts of 15 capacity from Centralia. Maybe if I could check a number 16 real quick. Again, it depends on the measure that you're 17 using. The energy that we're receiving out of Centralia 18 is 138 average megawatts compared to the load of roughly 19 950. It is in the neighborhood of 12 percent, yes. 20 Q Did I understand from a discussion earlier 21 that with respect to personnel within the resource 22 optimization department and energy trading and marketing 23 operations of Avista Energy there are common personnel? 24 A There are not common personnel between the 25 resource optimization department within the Company and 530 CSB REPORTING NORWOOD (X) Wilder, Idaho 83676 Avista 1 Avista Energy. They're operated independently. There 2 are no common traders or common support people between 3 the two entities. They're operated completely 4 separately, nor does Avista Energy rely upon or use in 5 any way the Company's assets. 6 Q Okay, it's as if there's a firewall there? 7 A There certainly is a firewall. At the 8 upper levels, obviously, there's some oversight that 9 occurs between the two entities. 10 MR. WOODBURY: Thank you. Madam Chair, I 11 have no further questions of Mr. Norwood. 12 COMMISSIONER SMITH: Do we have questions 13 from the Commission? Commissioner Kjellander. 14 15 EXAMINATION 16 17 BY COMMISSIONER KJELLANDER: 18 Q Mr. Norwood, earlier you were engaged in 19 some questioning regarding commercial trading and you 20 alluded to some of the risk limits that were in place and 21 I was just wondering if you could illuminate a little 22 further on what those actuals risk limits are. 23 A Yes. There's a number of provisions or 24 policies, procedures that are in place that have been put 25 together by the Company's risk management committee. 531 CSB REPORTING NORWOOD (Com) Wilder, Idaho 83676 Avista 1 There is a risk manager within the resource optimization 2 group that oversees that. One limit is the open 3 positions that the Company is allowed to have at any 4 point in time. For example, over the next three months 5 for the Company they're not allowed to have any more than 6 300 megawatts of open positions for those three months. 7 As you get beyond that, for the next year 8 there are also limits on the amount of open positions. 9 What I mean by open positions is if you buy power for 10 trading purposes, you can only buy so much before you run 11 up against that limit, you can't buy any more to have 12 open positions that are exposed to changes in prices. 13 We have a term called value at risk or VAR 14 that's calculated on a daily basis which measures the 15 exposure to the Company from a one-day move in market 16 prices and that limit is set at 1.5 million for the 17 Company. We also do stress tests on the open positions 18 and actually on the maximum open positions to ensure that 19 if there's a substantial move in prices on the open 20 positions that it doesn't create a substantial amount of 21 exposure to the Company. 22 We have, for the utility itself, we have a 23 limited number of partners that we trade with to minimize 24 the exposure to credit for parties that don't pay their 25 bills. We have a -- there is a credit function within 532 CSB REPORTING NORWOOD (Com) Wilder, Idaho 83676 Avista 1 the group which evaluates the corporate credit of those 2 parties that we do do business with to ensure that we can 3 be assured that we will be paid for those sales that we 4 make to them. 5 We have netting agreements in place with 6 the counter parties so that to the extent that other 7 parties owe us money and we owe them money, we can net 8 the two together so that we're not exposed to the full 9 amount. There's a limited number of products that this 10 group can trade. They can trade the futures to some 11 degree, but there are limits on the futures contracts 12 that they can trade. They can trade forward contracts 13 with a limited amount of options that they can trade. 14 They cannot do the elaborate transactions that you would 15 see at many of the trading companies, so if you look at 16 it in total, there's a limit on total positions they can 17 take, the products that they trade, the parties they 18 trade with and total value outstanding, so there's some 19 specific limits that are set for the Company to minimize 20 the exposure to the Company from this type of activity. 21 Q I hate to keep you up here, I realize 22 you've been here for a long time, but you mentioned the 23 value at risk was 1.5 million. I didn't quite catch the 24 time frame for that. 25 A That's a standard calculation in the 533 CSB REPORTING NORWOOD (Com) Wilder, Idaho 83676 Avista 1 industry with regard to trading activities. It's for a 2 one-day move in price. That's just a measure that they 3 use and some like it, some don't like it. It's just one 4 measure to evaluate risk. 5 COMMISSIONER KJELLANDER: Thank you very 6 much. 7 8 EXAMINATION 9 10 BY COMMISSIONER SMITH: 11 Q Mr. Norwood, in your discussion with 12 Mr. Ward on this allocation of costs which he called 13 incremental, I guess it caused me to wonder, the volume 14 of your commercial or speculative transactions exceeds 15 that done for your system load, doesn't it? 16 A Yes, it does. 17 Q So is it time, then, to just do an 18 incremental cost allocation to those who are doing system 19 load transactions instead of the other way around? 20 A Hopefully, either way you do the 21 calculation you'd end up with the same result. 22 Q Do you really believe that? 23 A I do. We took a careful look at the people 24 in the group that dedicate any part of their time to the 25 commercial trading activity and we included the risk 534 CSB REPORTING NORWOOD (Com) Wilder, Idaho 83676 Avista 1 manager and the director of corporate credit, other 2 people in the finance department, that's all picked up. 3 Q Then I guess I have to ask, why are the 4 employees who are doing the commercial transactions 5 apparently so much more efficient than the employees who 6 are doing the system transactions? 7 A It's not a matter of efficiency. It's a 8 matter of the function that they're entering into and I 9 was going to look in my testimony just as an example of 10 one transaction. It just takes a matter of -- for the 11 trading activity, it takes a matter of seconds for a 12 trader to call a broker and say what's the price and then 13 to buy 10,000 megawatt-hours and in this case, it was 14 30,000 megawatt-hours at $36.00 a megawatt-hour. That's 15 over a million dollars in revenue to sell one transaction 16 and then when they buy that back, again, that's over a 17 million dollars of purchase costs associated with that 18 one transaction, and so in a very short period of time 19 you can have huge volumes of dollars and megawatt-hours 20 in a very short period of time. 21 Q Well, Mr. Matthews and I discussed this 22 morning whether you keep an integrated utility or whether 23 you make them separate, would their regulated customers 24 be better off if we made them separate if the cost of 25 doing these transactions is so much less? 535 CSB REPORTING NORWOOD (Com) Wilder, Idaho 83676 Avista 1 A It's important to distinguish the types of 2 transactions they're entering into. As I said before, 3 the transactions they're entering into are a financial 4 firm product. There's no guarantees to deliver the 5 energy, nor is there a guarantee that the other party 6 will deliver to you. It's different than the type of 7 product that's being purchased to serve load. It's 8 different than managing resources to serve load and so 9 the time necessary to do the two different activities is 10 substantially different. 11 Q Okay, I guess I may have been misled by 12 those tickets we looked at where it said it was firm and 13 on peak. 14 A It is firm and on peak, but the players in 15 the industry are very concerned about what those products 16 represent and that's the reason in the WSPP agreement 17 they specifically laid out whether it's a financial firm 18 product or a system firm product, and we actually had 19 problems with this last year where a party that was 20 selling us power said that they were selling us "firm 21 power," but when it came down to it they said, well, I'm 22 not going to deliver, I'll just pay you liquidated 23 damages, you go buy it, so in the WSPP agreement, they 24 chose to specify, then, are you delivering firm power 25 backed by resources or are you selling financial firm 536 CSB REPORTING NORWOOD (Com) Wilder, Idaho 83676 Avista 1 power where there's no obligation to deliver but an 2 obligation on liquidated damages, so there are two 3 different products there. 4 Q Okay. I also wondered because it seemed 5 like the test year transactions were a lot greater than 6 net system load; that's correct? 7 A That is correct. 8 Q Then I'm wondering, it seemed like the 9 power supply model that you ran came up with normal 10 values that were just a small fraction of the actual 11 megawatt-hours. Is it time to look at changes in the 12 model? 13 A It depends on what the objective of the 14 model is and the objective of the model, what the model 15 has been used in this case as well as prior cases is to 16 model the cost associated with serving customers and the 17 wholesale obligations based on known contracts and loads 18 that are there and the existing resources. What isn't 19 picked up in the model and what isn't projected or 20 calculated in the model are those future commercial 21 trading activities or speculative transactions that you 22 can choose or not choose to enter into, but the dispatch 23 model does do what it has in the past, it does do what it 24 should do in this case and that is operate, figure out 25 what the costs are associated with operating the 537 CSB REPORTING NORWOOD (Com) Wilder, Idaho 83676 Avista 1 resources to serve retail load and firm contract 2 obligations. 3 The commercial trading activity is really a 4 separate issue of should the Company engage in those 5 activities and, if so, how much and it really is a 6 separate activity and based on the ratemaking that's in 7 place, to the extent they choose to do it, it really 8 falls to the bottom line whether they win or lose on 9 those transactions. 10 COMMISSIONER SMITH: Thank you. 11 Do you have redirect, Mr. Meyer? 12 MR. MEYER: I just have a few brief areas 13 just as a follow-on to your questioning. 14 COMMISSIONER SMITH: Excuse me, I think 15 Commission Kjellander has another question. 16 COMMISSIONER KJELLANDER: I do have one 17 more question and I know if I don't ask it, I'm going to 18 wake up in the middle of the night, which I'm going to do 19 anyway. 20 21 EXAMINATION 22 23 BY COMMISSIONER KJELLANDER: 24 Q Earlier during some of the testimony you 25 were providing where you showed some sample test years to 538 CSB REPORTING NORWOOD (Com) Wilder, Idaho 83676 Avista 1 demonstrate that you did not choose the highest value 2 period that you did your estimate on, I was curious, the 3 years 1934 to 1948 appeared in all of those other 4 examples, that 14-year time span and I think you also 5 mentioned within your testimony that you were very 6 familiar through that time period which ones were dry and 7 which ones were wet. That 14-year time span that occurs 8 in all those other examples that were present, what were 9 those 14 years like? 10 A Well, I guess all I can say about them is 11 that when you look at the actual streamflows that 12 occurred during that period based on the data that's 13 available was that there was a very dry period that 14 occurred during that time period. That's all I can tell 15 you is that it was a very dry time period back during 16 those years. 17 Q Specifically, then, with those 14 years, 18 that would have caught that very dry period, then? 19 A Well, I'm not sure if it's 14 as far as how 20 many years were involved during that dry time period. 21 There are some comparable years in the '80s that were 22 similar to the 1930s period. There weren't as many of 23 them, but there were certainly some very bad water years, 24 multiple bad water years, in the '80s that were very 25 similar to the '30s. As I mentioned before, if you look 539 CSB REPORTING NORWOOD (Com) Wilder, Idaho 83676 Avista 1 at other regional entities like Bonneville, they capture 2 that 1928 to '88 period to capture all that data because 3 it's recognized that kind of thing could reoccur. 4 COMMISSIONER KJELLANDER: Thank you. 5 6 EXAMINATION 7 8 BY COMMISSIONER SMITH: 9 Q Let me ask, when will they update it to 10 include the years we've had since '88? 11 A They update the study every 10 years and 12 the reason they wait 10 years is because it takes a lot 13 of time to go back through all the flows from all the 14 different tributaries and adjust those to reflect changes 15 in irrigation from farms and other depletion that's taken 16 place, so they go back and basically redo what the flows 17 would be based on current levels of irrigation. 18 Q So are they doing it now? It's been 10 19 years. 20 A It usually takes two or three years or more 21 beyond the end of the next 10-year period. 22 Q Which was '98. 23 A Which would go through '98. It will 24 probably be a number of years before we get that. In 25 1993, I think as I mentioned, was the time period that 540 CSB REPORTING NORWOOD (Com) Wilder, Idaho 83676 Avista 1 they -- that's when they brought out the next set of 2 10-year data which was five years following the 1988 time 3 period. 4 COMMISSIONER SMITH: Don't they have better 5 computers now? Mr. Meyer. 6 MR. MEYER: Fair enough. 7 8 REDIRECT EXAMINATION 9 10 BY MR. MEYER: 11 Q There's been a lot of discussion throughout 12 the afternoon about what you characterized or others have 13 characterized as commercial or speculative short-term 14 trading and you've testified as to how for ratemaking 15 purposes the Company has isolated the risk and the 16 benefit associated with that short-term trading. Do you 17 recall that testimony? 18 A Yes. 19 Q And would you distinguish again for the 20 benefit of the record and for the Commissioners the 21 difference between how we have treated the short-term 22 trading, any risks and losses associated with that, as 23 opposed to long-term marketing transactions that are not 24 trading transactions, per se? 25 A Right. There are a number in addition to 541 CSB REPORTING NORWOOD (Di) Wilder, Idaho 83676 Avista 1 the short term, there's the long-term transactions 2 which -- and this morning the term leverage was used, 3 leveraging the assets and our long-term group does 4 leverage the assets and engage in long-term transactions, 5 marketing transactions, which create substantial value 6 for the Company, and all of the value from all of those 7 long-term transactions has been included in this rate 8 case and has been in the past, so customers receive the 9 full benefit of all of those long-term marketing 10 transactions. 11 Q And by marketing transactions, do you mean 12 transactions that I guess in a colloquial sense touch the 13 system or make use in some fashion of the Company 14 resources? 15 A Most of them do. There have been some in 16 the past that do not, but the majority of those types of 17 transactions do, but many of them fall in the 18 neighborhood of some pretty creative transactions in 19 order to capture value either to increase margins or to 20 reduce costs. 21 Q There were some fairly significant dollars 22 discussed associated with short-term, you called them, 23 commercial transactions, others have characterized them 24 as speculative transactions. Do you recall that 25 exchange? 542 CSB REPORTING NORWOOD (Di) Wilder, Idaho 83676 Avista 1 A Yes. 2 Q Now, why is it not surprising that we have 3 seen that level of activity expressed in dollars on an 4 annual basis with that group of speculative transactions, 5 why is that not uncommon? 6 A Well, as I stated before, the margins, 7 maybe I haven't stated this, the margins are slim on 8 those types of transactions and it requires a large 9 volume of transactions in order to capture value and 10 again within the risk management parameters that are set 11 up there is a large volume that goes through a company's 12 books in the form of purchases and sales, so what you end 13 up with, then, is a large number in the way of short-term 14 purchases and short-term sales. 15 Q Now, switch back, if you will, to the 16 longer-term marketing transactions. You, of course, have 17 read the Staff prefiled testimony, haven't you? 18 A Yes. 19 Q And I believe there were estimates set 20 forth in Mr. Lobb's testimony suggesting that there were 21 substantial margins associated with those? 22 A Yes. He mentioned that there's a -- he 23 estimates the margins for customers from long-term 24 transactions of approximately $5.6 million per year. 25 Q And in the Company filing, do those margins 543 CSB REPORTING NORWOOD (Di) Wilder, Idaho 83676 Avista 1 inure solely to the benefit of ratepayers? 2 A Yes, the full amount of those margins have 3 been included for customers in this case. 4 MR. MEYER: May I approach the witness? 5 COMMISSIONER SMITH: Certainly. 6 (Mr. Meyer approached the witness.) 7 Q BY MR. MEYER: Mr. Ward earlier asked you 8 to review excerpted pages from Mr. Falkner's exhibits, 9 Exhibit No. 11? 10 A Yes. 11 Q And in an effort to put the power supply 12 adjustment that you're speaking to in this hearing in 13 some perspective, would you first turn to page 1 of 8 of 14 Exhibit 11 and specifically refer to line 23? 15 A I'm there. 16 Q Does that line reflect total pro forma 17 adjustments to the Company's case? 18 A Yes, it does. 19 Q What is the figure? 20 A It shows total adjustments of $51,845,000. 21 Q In round terms 52 million? 22 A Yes. 23 Q Now, would you turn lastly to page 7 of 8 24 of that same exhibit? 25 A I'm there. 544 CSB REPORTING NORWOOD (Di) Wilder, Idaho 83676 Avista 1 Q Line 24, please. 2 A Yes. 3 Q What do you show for your power supply 4 adjustment, pro forma adjustment? 5 A This show 15,258,000, so roughly $15 6 million for power supply. 7 Q 15 million out of approximately 52 million 8 in total pro forma adjustments? 9 A Yes. 10 Q Representing approximately a 29 percent 11 total pro forma adjustment? 12 A I'll accept that number. 13 Q Okay. Let's move to a series of 14 questioning that Mr. Woodbury had. He asked you about 15 Centralia, asked you about the sale of that, asked you 16 about the possible impact of that sale on items such as 17 fuel expense and other items. Part of your response 18 really expanded, I guess, the universe of factors that 19 would be affected. You talked about changes in O&M, in 20 capital, in transmission costs, you talked about 21 replacement power issues, did you not? 22 A Yes. 23 Q Would you agree with me that even with that 24 partial list of items or elements that would be impacted 25 by the sale of Centralia that there's substantial 545 CSB REPORTING NORWOOD (Di) Wilder, Idaho 83676 Avista 1 uncertainty? 2 A There's definitely a substantial amount of 3 uncertainty at this time. It's not a given that the sale 4 will go through. As we get closer to knowing whether the 5 sale will go through, we'll know more about what the 6 replacement power will be, what all the other costs will 7 be, so at this point in time there's a lot of uncertainty 8 as to whether the deal will go through and so it doesn't 9 make sense at this point in time to try to reflect 10 changes in this case, especially, because they're simply 11 not known or measurable. 12 Q Nor do you take issue, I gather, with 13 Mr. Matthews' earlier testimony today concerning the 14 probability of the sale or the timing of the sale in the 15 next 12 to 18 months? 16 A Correct. 17 Q Now, you're the sponsor of, as we've 18 discussed ad nauseam today, a variety of pro forma 19 adjustments that seek to bring into the power supply 20 adjustment certain known and measurable changes out 21 through June of 2000; correct? 22 A Yes. 23 Q And Mr. Ward spent some time with you on 24 page 4 of your direct testimony going through several 25 contracts which you deemed known and measurable. 546 CSB REPORTING NORWOOD (Di) Wilder, Idaho 83676 Avista 1 A Yes. 2 Q Contract changes; right? 3 A Yes. 4 Q And those were the sort of known and 5 measurable changes that you were sponsoring; correct? 6 A That's correct. 7 Q Now, compare and contrast those types of 8 pro forma changes, if you will, with the difficulty of 9 trying to pro form in at this time in this case all 10 impacts associated with the sale of Centralia. 11 A It would not be possible at this time to 12 pro form in all the impacts associated with Centralia 13 because there are too many factors that are unknown, too 14 many changes in costs that will occur if the sale goes 15 through that are unknown at this time, as I mentioned 16 before, the transmission costs, the replacement power 17 costs and so on; whereas, the contracts that I've 18 reflected and the changes I've reflected in this case are 19 based on known contract changes in general and so there's 20 a significant level of certainty associated with the 21 changes I've included in this case; whereas, there's a 22 significant level of uncertainty associated with the 23 Centralia project, and as we get closer to that sale 24 occurring, assuming that it does occur, we'll know more 25 and then there will be opportunity through a formal or 547 CSB REPORTING NORWOOD (Di) Wilder, Idaho 83676 Avista 1 informal filing at that time to then identify the changes 2 in costs associated with that particular transaction. 3 MR. MEYER: Thank you. That completes my 4 redirect. 5 (The witness left the stand.) 6 COMMISSIONER SMITH: I think that brings us 7 to the time today that we'll need to quit. We have a 8 public hearing scheduled for 7:00 p.m. tonight and I 9 think we all need to have a dinner break. I've looked at 10 the list of witnesses and based on how far we got today, 11 we only have three more days to go, so I think in the 12 morning we should plan to start about 8:30, so those of 13 you who need to be here at the public hearing, we'll see 14 you at 7:00 tonight. If not, we'll see you at 8:30 in 15 the morning. 16 (The Hearing recessed at 5:00 p.m.) 17 18 19 20 21 22 23 24 25 548 CSB REPORTING COLLOQUY Wilder, Idaho 83676