This is an ad valorem tax case involving the electrical generating facilities of Pacificorp, a Maine corporation doing business as Pacific Power & Light Company (Pacific). The evidence, including testimony from several expert witnesses and over one hundred exhibits, is extensive and complex. The Tax Court found the appropriate true cash value of Pacific for Oregon ad valorem tax assessment purposes to be $2,545,951,100. PP&L v. Dept. of Rev., 10 OTR 417 (1987). From that determination both Pacific and the Department of Revenue (Department) appeal. On de novo review, see ORS 305.445,19.125, we find that the true cash value of Pacific is $2,611,385,650.
The evidence suggests that Pacific is one of the most diversified electric utilities in the United States. It engages not only in electrical generation in six northwest states but also in coal mining, telecommunications, and other ventures. The electrical generation activity, which constitutes roughly half of Pacific’s business, is closely regulated by the public utility commissions of each of the six states in which Pacific does business as well as by the Federal Energy Regulatory Commission (FERC).
Ad valorem taxes are levied on the true cash value of the Oregon property of Pacific. ORS 308.515(l)(a); 308.540. Because Pacific’s electricity generation and distribution is integrated throughout its six-state service area, Pacific’s entire system is first valued as a unit, a portion of which then is allocated to Oregon. ORS 308.550; 308.555; see also United Telephone Co. v. Dept. of Rev., 307 Or 428, 430, 770 P2d 43 (1989) (describing similar process for valuation of assets of telecommunications company).
By regulation, Pacific’s property is appraised using three different approaches to valuation. See OAR 150-308.205-A. The results then are weighted to produce a final, composite valuation. These three approaches are the comparable sales approach, the cost approach, and the income approach. OAR 150-308.205(2); see also United Telephone Co. v. Dept. of Rev., supra, 307 Or at 432. Much of the record in this case is taken up with testimony concerning the way the parties’ appraisal experts carried out their valuation tasks under each of these approaches, including critiques by each *52side of the other’s methodology, analysis, and results. Many of the issues involved at that trial stage survive in this appeal. Those issues will be examined below in connection with our discussion of each of the three approaches to value.
Before we can deal with those questions, however, two preliminary matters must be considered. These involve establishing a brief glossary of terms pertinent to the balance of the opinion and answering certain procedural arguments advanced by the parties. We deal first with the glossary.
A. Glossary of Terms1
As a general proposition, electric utilities hold franchises from states to provide electric service to specified areas within the states. Within these areas, the utilities have a monopoly. In return for the monopoly, the utilities are permitted to charge for their products no more than is authorized by the state regulators. The regulators establish rates sufficiently high to permit investors in the utilities to realize a reasonable return on their investment, i.e., a return sufficient to attract continued investment in the utilities. The return is calculated by multiplying the value of the portion of a utility’s property that is devoted to providing electric service (a figure called the “rate base”) by a figure called the “rate of return,” expressed as a percentage. The utility is permitted to earn its return only on property devoted to providing electric service and, in certain cases, not even all of that property (as we will explain below).
Utilities must file periodic reports with their regulators detailing the property owned or leased by the utility and the purposes to which that property is put. ORS 308.520; 308.525. Commonly, such reports will show that a utility owns property that it is not currently devoting to electric service. In addition, special regulatory accounting is required for the following types of property:
1. “Zero capital cost” property. A basic premise of regulatory philosophy is that utilities are allowed to earn a reasonable rate of return on invested capital. If property has no capital cost, however, no return is allowed. There are three *53types of properties pertinent to this case that are treated by the regulators as having no capital cost. First, there is property purchased with funds held for “deferred income taxes” (DIT). Federal tax laws permit depreciation of certain kinds of property to be reported at a rate accelerated more than that actually experienced by such property. Thus, the utility creates an account for DIT. The savings may be invested in many ways, including in property to produce and transmit electricity. From the point of view of the regulators, however, property purchased in this way actually costs the utility nothing, so no return needs to be earned on that property.
A second tax-related phenomenon treated the same way as is DIT for regulatory purposes is investment tax credits (ITC). These are credits given under the Internal Revenue Code for the purchase of certain kinds of property for business use. Where funds derived from such credits are used to purchase electrical generating property, regulators again keep such property out of the rate base for the same reason property purchased with DIT is excluded.
The third category of zero capital cost property is known as “contributions in aid of construction” (CIAC). Pacific is required to furnish service to any customer within its service area. But where, for example, putting in new lines to a remote customer would cost more than any conceivable income to be derived from the service provided, the customer may be required to pay for a portion of the capital cost associated with bringing the service to the customer. The regulators do not allow Pacific to earn a return on that portion of the capital cost paid by the customer.
2. “Property not in service. ” A second basic premise of utility regulation is that a utility should be permitted to earn a return only on property that is reasonably necessary to and actually providing utility service. See ORS 757.355.2 The largest type of property in the property-not-in-service category is construction work in progress (CWIP). When a utility constructs new property, such as a generating facility, that *54property is not included in the utility’s rate base until it actually is placed in service and, even then, the regulators may not allow it in the rate base until the utility establishes that the property is reasonably necessary to provision of electrical service.
Another significant type of property in this category is property held for future use (PHFU). This is property — usually unimproved realty — which the utility anticipates it will need in the future but which it is not presently using to provide electric service. As long as such property is not used, the regulators will not allow the utility to earn a return on it.
3. “Leased property.” This category is the most straightforward. Because leased property used to provide electrical service is not owned by the utility, it is not included in the rate base. (Instead, the lease expense is treated as an allowable operating expense that should be recovered under the utility’s rate of return.) Nonetheless, ORS 308.517(1) specifically provides that such leased property shall be assessed “to the property user,” i.e., to the utility, for ad valorem tax purposes. Thus, the property is treated as the utility’s for tax purposes although it is not so treated for regulatory purposes.
All the foregoing terms play a role in the parties’ — and this court’s — analysis of the true cash value of Pacific. We turn next to the parties’ procedural arguments.
B. Procedural Arguments
Pacific contends that this court should disregard the appraisal submitted by the Department’s expert, Mr. Roger Maude, because Maude failed to specify how much weight he would give to his valuation under each of the three required valuation approaches in formulating his final, composite valuation. We reject this contention. Maude’s failure to indicate what weight he gave each approach may cause us to disregard his ultimate composite valuation, but it does not affect the relevance of the analysis he conducted or the results he obtained under each of the three approaches.
Pacific next argues that the Tax Court erred in not requiring the Department to carry the burden of proving that its appraisal, which suggested a value significantly greater than that found by the Director of the Department of Revenue *55at an earlier, administrative step in this case, should be adopted. Pacific cites no binding authority for this proposition, which contradicts the basic idea that the burden in an appeal by a taxpayer to the Tax Court is on the taxpayer. ORS 305.427. The Tax Court did not err. To the extent that Pacific is also asking that this court impose such a burden on the Department in this court, the request is denied. The burden is the same “upon appeal” from the Tax Court. ORS 305.427.
Finally, Pacific argues that the Tax Court failed to give appropriate precedential weight to prior decisions of this court concerning appropriate methods of valuing privately owned utility property. We considered and rejected precisely the same contention in United Telephone Co. v. Dept. of Rev., supra, 307 Or at 431-32. The decisions upon which Pacific relies, and which this court fully considered in United Telephone, did not establish the rule or rules of law Pacific now claims for them. They involved explanations of how this court, as a trier of fact on de novo review, found the facts in particular cases. The Tax Court properly rejected this contention. None of Pacific’s procedural arguments is well taken.
For its part, the Department argues that the Tax Court erred in denying its motions to reconsider and to reopen the record with respect to a particular issue under the cost approach to valuation. We find no error in the Tax Court’s discretionary refusal to reopen the proceedings. To the extent that the Department’s arguments also touch on the merits of the Tax Court’s factual determination of value under the cost approach, we shall deal with them under that heading below. The Department’s procedural argument is not well taken. We turn to the merits.
C. The Comparable Sales Approach to Valuation
While there was some dispute between the parties in the Tax Court concerning this approach, the dispute has not survived in this appeal. The parties recognize that, strictly speaking, it is not possible to value a large, integrated, privately owned electric utility like Pacific by the use of sales of comparable companies because comparable companies are not sold. Both Maude and Pacific’s expert, Dr. John Davis, offered a substitute technique, the “stock and debt *56approach.”3 The Tax Court accepted Davis’s result, viz., $2,539,000,000. The evidence establishes that this figure is reasonable, and the Department does not challenge it before us. We find that the value of Pacific as indicated by the stock and debt approach is $2,539,000,000.
D. The Cost Approach to Valuation
Davis found that Pacific’s value under the cost approach was $2,429,000,000. Maude found it to be $2,583,368,000. Both parties adopted Pacific’s historic cost less depreciation (HCLD), derived from Pacific’s annual report to the FERC, as the basis for their calculations of value under the cost approach. The difference arose primarily as a result of a deduction that Davis made for the amount of DIT and ITC. Davis called these two accounting entries a measure of “obsolescence,” because they represented an amount of plant that Pacific had but was not allowed to earn on — a situation essentially equivalent to having an inefficient plant that can only utilize a portion of its capacity to produce goods.
The Tax Court rejected the “obsolescence” label as “misleading” but nonetheless agreed that there should be deductions for both DIT and ITC because Pacific was not able to earn on them. The Tax Court explained:
“In the court’s view, the overriding principle or test is market value. If the market would not pay for property because the PUC will not allow that property to earn a return, no taxable value can be assessed to the regulated owner. The Department’s position [to the contrary] does not comport with reality. A purchaser would not pay the full price for the unit on a cost basis when the amounts for DIT and ITC will never be allowed to earn a return. By comparison, property in the CWIP and PHFU categories do have market value because they can become part of the recognized earning rate base. A buyer would undoubtedly discount the value of those properties from their cost based upon anticipated delays and perceived risks in getting them operating and into rate base. However, because of regulation, DIT and ITC property will not become part of the rate base. The court can see no reason *57why a buyer would pay cost for assets upon which he cannot earn a return.
“It is true that the tangible property represented by the amounts in DIT and ITC are part of the plaintiffs operational system. Those properties may in fact produce or deliver electricity. However, if they have any market value, it will have to be reflected as a part of the total system value when measured by the other approaches to value. It is probably misleading to speak of these circumstances as a form of ‘obsolescence.’ It would be more accurate to view such property as not having any value in the marketplace because government regulation has shifted the economic benefit to the public.”
PP&L v. Dept. of Rev., supra, 10 OTR at 427 (emphasis in original). The Tax Court therefore excluded DIT and ITC from HCLD. It found, consistent with Davis’s testimony, that Pacific’s value under the cost approach was $2,433,053,000. Id. at 428.
The Department argues that this analysis results in a portion of the tangible assets of Pacific inappropriately being exempted from taxation. We agree. Accepting (as the Tax Court and the parties apparently did) the idea that what we are searching for is what a hypothetical willing buyer of this property would pay to a hypothetical willing seller, it seems clear to us that a willing buyer of the plant and equipment would be agreeable to paying a figure close to HCLD, because the buyer could then earn off all that expenditure.
We say “close,” because there is much evidence in this record to the effect that regulation to some extent diminishes the earning potential of regulated property and, therefore, a willing buyer would wish to discount the property to some degree. Thus, Pacific’s property may be affected by a measure of obsolescence. But — as the Tax Court recognized —DIT and ITC are no measure of (or surrogate for) that obsolescence. Both are property purchased through tax allowances for certain capital investments.
Both DIT and ITC are property and have intrinsic value. Even Davis recognizes this. His exclusion of both items from HCLD is an attempt to find some surrogate for the obsolescence he insists is present, not a denial that they have value. Absent any satisfactory evidence establishing an appropriate measure of obsolescence, it is inappropriate to make *58any deduction for it by subtracting in toto the values for DIT and ITC — the only deduction for which Pacific argues. We agree with the Department that the Tax Court erred in making the deduction. Adding the deduction to the Tax Court’s finding, we find that the value of Pacific under the cost approach is $2,578,463,000.
E. The Income Approach to Valuation
Income approach theory assumes that an income-producing property is worth the present value of its future income stream. Mathematically under this approach, the value of a company is derived by use of the accounting formula, V = I/R. The “I” in the formula is an estimated future income figure for the company. The “R” is a capitalization rate that is divided into the income figure to obtain the value, “V.” The evidence shows that this approach is considered by appraisers to be quite reliable. There are, however, a number of philosophical and practical disagreements between the parties under this approach.
Davis found Pacific’s value under the income approach to be $1,920,000,000. Maude used three different calculation methods under this approach, resulting in valuations ranging from $2,346,910,000 to $2,674,041,000. The Tax Court valued Pacific under this approach at $2,660,394,000 — a figure due principally to the court’s acceptance of Maude’s methodology over Davis’s. For convenience, we shall follow the Tax Court’s analytical approach through the various steps of the process, commencing with the determination of the appropriate income figure (“I”).
1. Income (“I”). Davis derived his income figure by starting with Pacific’s net operating income (NOI) figures for the past five years and then averaging them in various ways in order to obtain an appropriate figure for projected income. The Tax Court rejected Davis’s income figure because it assumed or projected no “growth.” We reject it for the same reason. Pacific’s (then) most recent five-year experience showed growth; its own future projections expected growth. *59Growth in income was the most reasonable assumption to make.4
Rejection of Davis’s income approach left the Tax Court with Maude’s three methods for deriving a figure for “I,” all of which assumed growth. The court adopted, albeit with some adjustment, Maude’s “Income Adjusted” method.5 Under that method, future NOI is adjusted to reflect increased demand and the impact of inflation so that a correct figure for “growth” can be identified. Maude projected growth, after deducting for inflation, at 2.5 percent. The Tax Court rejected this assumption, for reasons explained in its opinion. Id. at 431-32. The court instead adopted a figure for growth of 4.5 percent based on testimony of Mr. Nyegaard, an employee of the Oregon PUC. Id. We do so, as well.
The Tax Court did not see fit to announce its calculation of a figure for “I” under its modification of Maude’s “Income Adjusted” approach. The court simply stated that its expected growth in income would be “at least equal to the growth anticipated by Mr. Maude.” Id. at 439. The court seems to have assumed that any difference would not need to be quantified, because it was adopting a capitalization rate lower than Maude’s and was accepting Maude’s figure for “I” — if there were any error, it would be in undervaluing Pacific.
The court’s failure to include its calculations was unfortunate for two reasons. First, including the figure for “I” and an explanation of the calculations that produced that figure would have permitted the parties to assure that no mathematical (as opposed to theoretical or evidentiary) errors were present. Second, knowing the figure might help this court in the event we accepted the Tax Court’s rationale as to *60“I” but believed that “R” should be larger than either the Tax Court or Maude calculated it to be. In such a case, the result might have been to lower the valuation of Pacific under this method. As it turns out, however, the court’s omission does no harm here, because we accept Maude’s figure for “R.” In the future, we hope that the Tax Court will see fit to explain fully its calculations in cases of this kind. We turn to an examination of “R.”
2. Capitalization Rate (“R”). The capitalization rate is the percentage return expressed as a decimal that an investor would expect to receive from the income stream in a given year. Our explanation of the capitalization rate in United Telephone Co. v. Dept. of Rev., supra, 307 Or at 435, is equally applicable here:
“Evidence from both experts [who were the same experts who testified in the present case] shows that different kinds of investment in the same company earn different rates of return. For example, the return an investor might receive on a long-term debt of a company would probably differ from the return the investor would receive through ownership of preferred stock, or of common stock. These three forms of investment are called ‘bands’ of investment by both experts. Because of these different rates of return, Davis and Maude agreed that it is appropriate to establish different capitalization rates for these ‘bands’ of investment, then weight each of the three bands according to the percentage of participation of each in the overall capital structure of the company, and take an average.”
The parties argue two issues related to the bands of investment — the appropriate capital structure (i.e., the percentage to be attributed to each band) and the appropriate cost of equity (i.e., the return an investor would expect for Pacific’s common stock). We shall discuss each at the appropriate point.
Both appraisers looked at capital structures of other electric utilities in order to determine what capital structure a prospective investor would expect. Their figures, together with Pacific’s actual capital structure, were as follows:
*61DAVIS MAUDE ACTUAL
Debt 48.3% 46.6% 46.86%
Preferred 11.4% 11.2% 14.01%
Equity (Common) 40.3% 42.2% 39.13%
The debt-to-equity ratio is important in two different respects. First, as the Tax Court noted,
“the overall cost of capital is less if equity is increased to more than 36 percent. [Citation to record omitted.] Thus selection of an appropriate capital structure will influence the appraiser’s selection of an appropriate weighted average cost of capital.”
PP&L v. Dept. of Rev., supra, 10 OTR at 433. Second, the larger the proportion of debt, the riskier the venture— creditors have to be paid regularly, while dividends to equity holders may be foregone.
Davis’s average capital structure was based on the average capital structure of the electric utility industry as a whole. Maude’s average capital structure was more tailored. It was drawn from a selection of 19 electric utilities whose bond ratings (as reported by an investment service) were similar to Pacific’s bond rating. The Tax Court found Maude’s approach to be the better one.
Pacific challenges the Tax Court’s choice of average capital structure on a variety of grounds, none of which persuades us that Davis’s approach is better. On the whole, we are slightly more persuaded that the structure proposed by Maude is the one an investor would use if the investor were evaluating a possible investment in Pacific. It is a more focused structure than the industry average, yet sufficiently diversified to seem reliable as a predictive device. We agree with the Tax Court that Maude’s average capital structure is the appropriate one to use.
Turning to the cost of capital, we note that Pacific challenges only the Tax Court’s finding as to the cost of equity capital. It does not challenge the court’s acceptance of Maude’s cost of long-term debt (13.0 percent) or his cost of preferred stock (12.9 percent). As these last figures are unchallenged, we shall adopt them without further analysis.
As for the cost of equity capital, the appraisers each tried out several different methods in seeking an appropriate *62figure. Davis put his primary emphasis on the “risk premium” method. Under this method, the appraiser first establishes a “risk free” equity rate — usually the rate currently charged for some United States Treasury security. To this the appraiser adds a “premium” — a figure representing the degree of risk associated with the particular company’s stock, compared with the risk of a broader range of companies of all types whose stocks have a significant trading history. The total of the risk-free and premium components equals the cost of equity capital. See United Telephone Co. v. Dept. of Rev., supra, 307 Or at 437-41 (explaining “risk premium” method and illustrating its application).
Davis took as his risk-free figure 8.94 percent, which was the interest rate for 30-year Treasury bonds as of December 31,1983. To this figure he added a risk premium of 8.30 percent to obtain an indicated equity rate of 17.24 percent.6 The Tax Court rejected Davis’s calculation on the ground that Davis had used inconsistent risk indicators for the “risk premium” method and for another method known as the “capital asset pricing model” (CAPM). Pacific argues that the Tax Court has assumed a connection between the risk analysis in the two methods that does not exist.
We agree with Pacific that the Tax Court at least used inappropriate terminology in explaining its rejection of Davis’s risk premium result. The court spoke of Davis using a “beta” in his risk premium calculation, but “beta” is a concept used in CAPM calculations, not in risk premium calculations. Terminology notwithstanding, however, we think the Tax Court’s underlying theory was correct. Davis’s 8.30 percent premium was a figure from Ibbotson-Sinquefield representing the average risk of the entire range of companies in the Standard & Poor 500 index. Many of those companies are riskier than Pacific, which is regarded as being relatively risk free due to the same condition Pacific generally decries: regulation. Davis’s calculation of the cost of equity capital is too high.
Other credible evidence suggested that Pacific had *63itself indicated to the Oregon PUC that a cost of equity capital of 14.6 percent was reasonable. Maude calculated the cost to be 15.2 percent. The Tax Court, persuaded that Davis’s was too high, accepted Maude’s figure. We agree that Davis has to be too high. Pacific does not argue for any figure between Davis and Maude. We, too, accept Maude’s 15.2 percent figure.
Having found all the figures necessary to calculate “R,” the Tax Court then selected .14 (i.e., 14 percent) as its weighted average cost of capital.7 It then stated Pacific’s value under this income adjusted model to be $2,269,240,000— precisely Maude’s figure, although the mathematics using .14 for “R” would not have worked out that way. As we have already noted, however, no one potentially aggrieved by the “R” figure or the resulting mathematical calculation now complains, so we accept the figure announced by the court.
As a final step under the income approach to valuation, the Tax Court added three figures representing adjustments that it deemed should be made to the income indicator of value — figures representing the capitalized income expected to be derived from CWIP, PHFU, and leased property. The court’s calculation was as follows:
Maude’s income adjusted model $2,269,240,000
Construction work in progress 213.893.000
Property held for future use 2,491,000
Leased property 174.770.000
Total income indicator $2,660,394,000
Of the additions made by the Tax Court, Pacific challenges only that for CWIP. It argues that the addition is inappropriate for two reasons: first, it is contrary to precedent; and, second, such an addition amounts to “double counting” of assets.
Pacific’s precedential argument is based on this court’s decision in a previous ad valorem tax case involving Pacific, Pacific Power & Light Co. v. Dept. of Revenue, 286 Or *64529, 549, 596 P2d 912 (1979), in which this court denied addition of CWIP that it found “too speculative.” That CWIP, however, involved significant outlays for nuclear power plants that this court doubted would ever reach the rate base — a doubt more than substantiated by subsequent events. The present CWIP is not so suspect. Thus, even if our previous decision as to this question were legal precedent — and, as we have already explained, it is not — the precedent would not be in point.
Secondarily, Pacific argues that the estimate for “I” in the equation V = I/R already assumed additional plant, because the “I” proposed by all appraisers was “trended up,” i.e., was higher than current income. As we have already indicated, however, we do not regard increase in plant as a sine qua non of higher income. We thus do not accept Pacific’s premise for this argument.8
We find Pacific’s value under the income approach to be $2,660,394,000.
The remaining task is to correlate the figures derived from the three approaches to value. Our only disagreement with the Tax Court occurred with respect to the cost approach — we found a value higher than that found by the Tax Court. Making that adjustment, the final calculation is as follows:
PROPORINDICATED TIONATE APPROACH VALUE WEIGHT VALUE
Cost $2,578,463,000 45%9 $1,160,308,350
Stock/Debt $2,539,000,000 10% 253,900,000
Income $2,660,394,000 45% 1,197,177,300
Total $2,611,385,650
We find that the true cash value of Pacific for ad valorem tax purposes as of January 1, 1984, was *65$2,611,385,650. The judgment of the Tax Court is modified accordingly.
The judgment of the Tax Court is modified. The case is remanded to the Tax Court for further proceedings consistent with this opinion.
Our glossary of terms is derived primarily from a portion of the opinion of the Tax Court in this case. See PP&L v. Dept. of Rev., 10 OTR 417, 421-23 (1987).
ORS 757.355 provides:
“No public utility shall, directly or indirectly, by any device, charge, demand, collect or receive from any customer rates which are derived from a rate base which includes within it any construction, building, installation or real or personal property not presently used for providing utility service to the customer.”
Use of the “stock and debt approach” as a substitute technique is authorized by OAR150-308.205-(A)(2)(d),andisdiscussedin United Telephone Co. v. Dept. of Rev., 307 Or 428, 432-33, 770 P2d 43 (1989).
In its second assignment of error, Pacific argues that Davis did, in fact, project growth — his projected income figure for 1984 ($270 million) was greater than Pacific’s NOI for 1983 ($250 million). But this figure appears to have been a mechanistic allowance for possible growth in the rate base, not a recognition that, among other things, demand might cause Pacific to utilize more fully its generating capacity to make more money.
The Tax Court did not adopt — indeed, it did not even discuss — Maude’s other two methods, viz., the “Rate Adjusted” method and the “Limited Life” method. Because we, like the Tax Court, find sufficient utility in Maude’s “Income Adjusted” method, we also do not utilize the other two methods. This moots Pacific’s sixth and seventh assignments of error, which challenged those methods.
The premium for risk was derived from the Ibbotson-Sinquefield Study, a comprehensive and authoritative study of various risks and returns associated with particular industries in relationship to the stock market as a whole. See United Telephone Co. v. Dept. of Rev., supra, 307 Or at 437 (discussing use of Ibbotson-Sinquefield Study).
The Tax Court acknowledged its figure “is higher than the calculated [weighted average capital cost] which would result from Mr. Maude’s capital structure with 42.2 percent equity.” Our own rough estimate puts “R” at closer to .1291. The Department does not challenge this determination, however, and we shall accept it.
Pacific also argues that the Tax Court incorrectly put the burden of proof on it to show that CWIP should not be included, rather than requiring the Department to show that it should. Because we think the record affirmatively justifies including CWIP, the placing of the burden of proof on the issue hardly matters.
No issue is raised at this stage as to the percentages chosen by the Tax Court for proportionate weighting of the various valuation figures.