Opinion
HILL, P. J.This is an appeal by Kern County from the trial court’s judgment, which rejected the decision of the Kern County Assessment Appeals Board (board) upholding the county tax assessor’s increased valuation of plaintiffs’ business property and the resulting increased property tax. The county contends the trial court applied the wrong standard in reviewing the administrative decision; it contends application of the correct standard would have resulted in a judgment upholding the administrative decision because it was supported by substantial evidence. The county further asserts that the trial court erred when it admitted new evidence that was not presented to the board and when it determined the tax assessor used incorrect revenue figures in calculating the income stream on which the property value was based in one of the appraisals. We conclude the trial court applied the correct standard of review, properly admitted evidence at trial, and correctly rejected the county’s revenue figures. Contrary to the trial court’s judgment, however, the matter must be remanded to the board for further proceedings because factual questions remain.
*725 FACTUAL AND PROCEDURAL BACKGROUND
Plaintiffs in these consolidated actions, Sky River LLC (Sky River) and Mojave 16/17/18 LLC (Mojave), are two related limited liability companies1 that own and operate wind farm electricity generation facilities in Kern County. They challenged the tax assessments of the Kern County tax assessor for plaintiffs’ business property2 for 2006 and 2007. The property consisted of wind turbine generators and related equipment, used to generate and transmit electricity. Both plaintiffs paid the taxes and initiated administrative proceedings before the board to challenge the valuations of their property. The board held two hearings; it reviewed the complex calculations the tax assessor used in computing the tax and the alternative calculations proposed by plaintiffs, and found in favor of the county. The board approved the tax assessor’s increased valuation of the property. Plaintiffs then filed actions in the superior court for a refund of a portion of the property taxes paid, again asserting that the tax assessor overvalued the property. The trial court found in favor of plaintiffs, concluding the tax assessor used a flawed methodology in calculating the value of the property, resulting in an inflated value and an overstated tax. It adopted the method of computation plaintiffs advocated and accepted the corrected figures they submitted. The trial court entered judgment setting out the corrected values for each plaintiff’s property and ordering a refund of any excess tax paid. The county appeals from the judgment of the trial court.
DISCUSSION
I. Standard of Review
The initial issue presented by this appeal concerns the appropriate standard of review in the trial court. The county contends the board’s decision was based on factual determinations; accordingly, the trial court should have deferred to the board’s findings of fact and should have reviewed the decision only to ascertain whether it was supported by substantial evidence. Plaintiffs contend the material facts were undisputed and the issue was whether the tax assessor applied the proper methodology in calculating the value of the property, in accordance with applicable statutes and regulations. Plaintiffs assert this was an issue of law, which the trial court properly reviewed de novo. The correct standard of review depends upon the nature of the dispute before the trial court.
*726Property subject to taxation must be assessed at its full value, which is defined as its full cash value or fair market value. (Rev. & Tax. Code, §§ 110.5, 401.) “ ‘[F]ull cash value’ or ‘fair market value’ means the amount of cash or its equivalent that property would bring if exposed for sale in the open market under conditions in which neither buyer nor seller could take advantage of the exigencies of the other, and both the buyer and the seller have knowledge of all of the uses and purposes to which the property is adapted and for which it is capable of being used, and of the enforceable restrictions upon those uses and purposes.” (Rev. & Tax. Code, § 110, subd. (a).) There are three basic methods for calculating fair market value: (1) the comparative sales or market data method; (2) the reproduction or replacement cost method; and (3) the income method. (Cal. Code Regs., tit. 18, §§ 3, 4, 6, 8; Pacific Mutual Life Ins. Co. v. County of Orange (1985) 187 Cal.App.3d 1141, 1147 [232 Cal.Rptr. 233].) In this case, the parties agree the value of the property should be calculated by using the income method, which is described in California Code of Regulations, title 18, section 8 (Rule 8).3
The income approach seeks to determine “[t]he amount that investors would be willing to pay for the right to receive the income that the property would be expected to yield, with the risks attendant upon its receipt.” (Cal. Code Regs., tit. 18, § 3, subd. (e).) “Using the income approach, an appraiser values an income property by computing the present worth of a future income stream. This present worth depends upon the size, shape, and duration of the estimated stream and upon the capitalization rate at which future income is discounted to its present worth.” (Rule 8, subd. (b).) “ ‘The income method rests upon the assumption that in an open market a willing buyer of the property would pay a willing seller an amount approximately equal to the present value of the future income to be derived from the property.’ [Citation.] ‘. . . “The income approach may be called the capitalization method because capitalizing is the process of converting an income stream into a capital sum, i.e., value.” [Citations.]’ ” (Freeport-McMoran Resource Partners v. County of Lake (1993) 12 Cal.App.4th 634, 642 [16 Cal.Rptr.2d 428] (Freeport).)
“ ‘The assessor capitalizes “the sum of anticipated future installments of net income from the property, less an allowance for interest and the risk of partial or no receipt.” [Citation.]’ [Citations.]” (Freeport, supra, 12 Cal.App.4th at p. 642.) “The discount factor or capitalization rate which is *727applied reflects interest, the risk of no return or a lesser return of income, liquidity, investment management, taxes, and depreciation, where appropriate. [Citation.] Thus, a high-risk investment carries a proportionately higher capitalization rate.” (Texaco Producing v. County of Kern (1998) 66 Cal.App.4th 1029, 1037 [78 Cal.Rptr.2d 433] (Texaco).) “Since a property’s ‘full value’ must be determined by reference to the price it would bring on an open market, ‘[t]he net earnings to be capitalized ... are not those of the present owner of the property, but those that would be anticipated by a prospective purchaser.’ ” (Freeport, at p. 642.)
There are two means by which a capitalization rate may be developed. (Rule 8, subd. (g).) “The preferred method is to derive the rate from the market by ‘comparing the net incomes that could reasonably have been anticipated from recently sold comparable properties with their sales prices, adjusted, if necessary, to cash equivalents .. . .’ [Citation.]” (Texaco, supra, 66 Cal.App.4th at p. 1037.) Alternatively, when there is a lack of comparable property sales, a “band-of-investment method” may be used: a discount rate is developed by “deriving a weighted average of the capitalization rates for debt and for equity capital appropriate to the California money markets.” (Rule 8, subd. (g)(2).) Using this method, also known as the weighted average cost of capital (WACC) method, the appraiser must “weight the rates for debt and equity capital by the respective amounts of such capital he deems most likely to be employed by prospective purchasers.” (Ibid.) “This method is based on the premise that die yield rate is the weighted average of the return on the different portions of the investment, i.e., debt and equity. [Citation.]” (Texaco, at p. 1038.)
The income stream to be capitalized pursuant to Rule 8 is “the net return which a reasonably well informed owner and reasonably well informed buyers may anticipate on the valuation date that the taxable property . . . will yield.” (Rule 8, subd. (c).) Net return is the difference between gross return and gross outgo; gross outgo does not include amortization, depreciation, property taxes, or income taxes. (Ibid.) Because taxes are not excluded from net return, the income stream to be capitalized is before tax. Accordingly, the discount rate used must also be before tax. The discount rate computed pursuant to the WACC method, however, is an after-tax rate, which must be adjusted to a before-tax rate in order to apply it to the income stream. (SBE, Assessors’ Handbook, Section 502, Advanced Appraisal (Dec. 1998) p. 98 & appen. A, p. 180 (SBE Assessors’ Handbook) <http://www.boe.ca.gov/proptaxes /pdf/ah502.pdf> [as of Feb. 21, 2013].) This adjustment is made “by dividing the after-tax WACC by 1 minus the combined state and federal corporate income tax rate.” (SBE Assessors’ Handbook, supra, appen. A, p. 180.) The income tax rate to be used is “the combined federal and state marginal corporate income tax rate (i.e., the tax rate paid on each incremental dollar of *728income).” (Id., at appen. A, p. 178; see id., at p. 98.) The higher the income tax rate used, the higher the discount rate and the lower the value of the property.
The parties agree' the income approach is the appropriate method of valuation, and due to a lack of sales of comparable property, the band-of-investment approach should be used to calculate the discount rate. Although the parties disagreed in the administrative proceeding about how this band-of-investment, or WACC, method would apply in calculating the discount rate, in the trial court plaintiffs accepted the method proposed by the assessor, with the exception of the income tax rate to be used to adjust the discount rate to a before-tax rate.
In this appeal, the parties again agree that the income method should be used to calculate the value of the property in issue. They also agree that the band-of-investment method should be used to develop the discount rate to be used to establish the present value of the future income stream. They disagree, however, on the income tax rate to be used in converting from an after-tax discount rate to a before-tax discount rate. The assessor calculated an average rate of corporate income tax he believed a potential purchaser of the property would anticipate. Plaintiffs contend the maximum combined federal and state marginal rate should be used.
Because of the parties’ agreement on certain aspects of the methodology to be used, the county contends the parties agree on the methodology and disagree only on the factual data to “plug in” to make the calculations. Because it contends only factual matters are in issue, the county argues that the substantial evidence standard of review applies. Plaintiffs, however, argue that the dispute relates to one element of the methodology—the appropriate income tax rate to apply to the conversion—rather than to factual matters, so the de novo standard of review applies. We review case authority for guidance.
In Maples v. Kern County Assessment Appeals Bd. (2002) 96 Cal.App.4th 1007 [117 Cal.Rptr.2d 663] (Maples), the owner of a low-income housing development challenged the assessor’s valuation of the property; under the federal program pursuant to which the housing development was operated, the owner paid market interest on its mortgage, but received a tax credit for the interest, which reduced the effective mortgage rate to 1 percent. The tax assessor valued the property using a cost approach; it bolstered its argument for that value by showing that an income approach to valuation, calculated using the 1 percent effective mortgage rate in the debt component of the band-of-investment capitalization rate calculation, reached a similar result. (Id. at pp. 1011-1012.) The owner contended an income approach should have been used, and the value should have been calculated using a market interest *729rate. (Id. at p. 1011.) The county appeals board agreed with the owner that the income approach should have been applied, using a market interest rate. (Id. at p. 1012.) The assessor sought a writ of mandate in the trial court; the trial court treated the issue as one of law and concluded the income method was the proper valuation method, but the 1 percent effective mortgage interest rate should have been used for the debt component of the band-of-investment capitalization rate calculation. (Ibid.)
On appeal, the parties disagreed on the standard of review: whether the appellate court was to defer to the board’s findings of fact and apply a substantial evidence standard or determine the issue de novo. (Maples, supra, 96 Cal.App.4th at pp. 1012-1013.) The court summarized the applicable standards of review: “Where the taxpayer claims a valid valuation method was improperly applied, the trial court is limited to reviewing the administrative record. [Citation.] The court may overturn the assessment appeals board’s decision only if there is no substantial evidence in the administrative record to support it. [Citation.] However, where the taxpayer challenges the validity of the valuation method itself, the court is faced with a question of law. In such a case, the court does not evaluate whether substantial evidence supports the board’s decision, but rather must inquire into whether the challenged valuation method is arbitrary, in excess of discretion, or in violation of the standards prescribed by law. [Citation.]” (Id. at p. 1013.) The court noted: “Whether a taxpayer is challenging ‘method’ or ‘application’ is not always easy to ascertain. [Citation.] If none of the facts are in dispute, what might otherwise appear to be a factual challenge, and therefore subject to substantial evidence review, is actually a legal challenge. [Citation.] ‘ “The issue is not whether the assessor misunderstood or distorted the available data, but whether he or she chose an appraisal method which by its nature was incapable of correctly estimating market value.” ’ [Citation.]” (Id. at pp: 1013-1014.) Determining that the only issue presented was whether to use the subsidized interest rate or a general market interest rate, the court concluded the question was one of law, reviewable by the trial court and the appellate court de novo. (Id. at p. 1014.)
In Bret Harte Inn, Inc. v. City and County of San Francisco (1976) 16 Cal.3d 14 [127 Cal.Rptr. 154, 544 P.2d 1354] (Bret Harte), the plaintiff, owner of a hotel, challenged the amount of its personal property tax assessment. The assessor had determined the value of the plaintiff’s personal property by taking the original cost and deducting 50 percent for depreciation, regardless of the age or condition of the particular items of property. (Id. at p. 18.) Undisputed evidence presented by the plaintiff at trial indicated its property had a lower value than that ascribed to it by the assessment. (Id. at p. 19.) The trial court entered judgment for the plaintiff, after concluding the valuation method was invalid, and the defendant appealed.
*730The court concluded that, when a taxpayer challenges the validity of a valuation method, it presents a question of law: whether the challenged method of valuation is arbitrary, in excess of discretion, or in violation of the standards prescribed by law. (Bret Harte, supra, 16 Cal.3d at p. 23.) The court treated the question before it as a question of law. It determined the cost method of valuation could be used, but only if it was “designed so that cost factors, which by their nature can have no direct relationship to present value, are modulated by depreciation factors in a manner reasonably calculated to achieve an approximation of such value with respect to the individual taxpayer.” (Id. at p. 25.) The method used by the assessor failed to do this; it discounted the original cost by a uniform 50 percent, regardless of age or condition of the property, without attempting to distinguish among individual properties and establish their current individual values. (Ibid.) This method was arbitrary, in excess of the assessor’s discretion, and in violation of the constitutional and statutory requirements that all property be assessed at its full value. (Ibid.)
Similarly, in ITT World Communications, Inc. v. County of Santa Clara (1980) 101 Cal.App.3d 246 [162 Cal.Rptr. 186] (ITT World), the SEE, which had operated under a rule that valuation of state-assessed utility property could not exceed its reproduction cost new less depreciation (RCNLD), rescinded that rule and subsequently assessed plaintiff’s property at an amount in excess of the RCNLD. The court concluded the issue raised by plaintiff—whether the method used by the board to assess its property was itself illegal because it did not retain the use of RCNLD as a ceiling on value—presented a question of law. Thus, the court was required to determine whether the board’s abandonment of the RCNLD as a ceiling “was arbitrary, in excess of discretion, or in violation of the standards prescribed by law.” (Id. at pp. 252-253.)
In Bontrager v. Siskiyou County Assessment Appeals Bd. (2002) 97 Cal.App.4th 325 [118 Cal.Rptr.2d 182] (Bontrager), the plaintiff owned rental housing subsidized by a federal government program; in exchange for the plaintiff charging lower rent to eligible tenants, the government subsidized the interest rate on financing for the property, so the plaintiff paid only 1 percent. The plaintiff challenged the tax assessor’s assessment of the value of the property. The parties agreed that the income approach and the band-of-investment method should be used to determine the property value. (Id. at p. 331.) The plaintiff argued, and the board agreed, that the market rate of interest should be used in determining the debt component of the WACC calculation. (Id. at p. 329.) The trial court disagreed. On appeal, the court held the correct interest rate to use was the actual rate paid by the plaintiff, i.e., 1 percent; use of the market rate would not produce an accurate estimate of the fair market value of the property. (Id. at p. 333.) The court rejected the plaintiff’s contention that the question was one of fact, subject to substantial *731evidence review. “[T]he determination of whether the face rate of interest charged by a lender or the actual amount of interest paid by the borrower should be used in calculating the capitalization rate is a question of law, both before the trial court and on appeal. . . . The factual issue, the amount of that rate, is undisputed.” (Id. at p. 334.)
Like the taxpayers in Maples, Bret Harte, ITT World, and Bontrager, plaintiffs in this case contend the assessor used an improper method to calculate the value of their property and the amount of their tax. Plaintiffs contend that, although the assessor correctly chose the income approach to determine value, and the band-of-investment method to calculate the capitalization rate, he improperly used an estimated average income tax rate, rather than the marginal rate prescribed by the assessor’s handbook, when converting the discount rate from an after-tax rate to a before-tax rate. We conclude that plaintiffs are correct, and the issue presented constitutes a question of law as to an element of the chosen method to be used in calculating the market value of the property. Which income tax rate should be used—the marginal rate or an average rate—is a question about the method of calculating the appropriate conversion rate. The exact percentage to be used for that rate would be a question of fact to be determined by the board based on the evidence presented. Determining which rate should be used does not present a question about the facts specific to plaintiffs’ case or the data to insert when calculating the value of the property. Rather, it presents a question about the methodology prescribed by SEE rules for calculation of the property value.
II. Proceedings Before the Board and at Trial
Because it is presumed that the assessor has properly performed his duties, the burden is on the taxpayer challenging an assessment to prove that the value on the assessment roll is not correct. (Cal. Code Regs., tit. 18, § 321, subd. (a).) This presumption, however, disappears when the assessor seeks to increase the valuation of the taxpayer’s property; in that event, the burden is on the tax assessor to prove the higher value of the property. (Id., § 313, subd. (f).) In this case, the assessor sought to raise the enrolled value of the properties, so the county bore the burden of proof of the increased value.
To determine the value of plaintiffs’ property using the income method, the assessor must compute the present worth of a future income stream. (Rule 8, subd. (b).) The future income stream is a forecast of the income expected to be generated by the property being appraised; it is the income that a typical, prudent buyer would expect the property to yield over the income projection period. (SEE Assessors’ Handbook, supra, at p. 56.) In his appraisals for 2006 and 2007, the assessor calculated an expected future income stream for Mojave’s property until 2028 and for Sky River’s property until 2020. The *732assessor also estimated expenses for the same periods, based on plaintiffs’ operating histories. He subtracted the expenses from the income to arrive at annual net income figures for the projection period. He then calculated the rate to use to discount that net income to present value by using the band-of-investment method.
The parties agree the appropriate method for developing a discount rate is the band-of-investment method. (Rule 8, subd. (g)(2); Maples, supra, 96 Cal.App.4th at p. 1011.) In this method, the capitalization rates attributable to components of a capital investment (debt and equity) are weighted and combined to derive a weighted-average rate attributable to the total investment. (SBE Assessors’ Handbook, supra, at p. 78.) “The first step in performing a band-of-investment analysis is to determine the percentage of debt that would be appropriate for the specific property. The percentage remaining after subtracting this debt from 100 is the equity. The debt percentage is then multiplied by the cost of the debt, i.e., the mortgage rate, and the equity percentage is multiplied by the expected rate of return for this investment. Adding these two figures together provides an overall indication of the discount rate.” (Texaco, supra, 66 Cal.App.4th at p. 1038.)
A variation of this WACC technique, called the pure play or comparable company method, uses data from the capital market relating to publicly traded companies engaged in a line of business similar to that of the property being valued to estimate a discount rate for the cash flow of an entity that is not publicly traded. (SBE Assessors’ Handbook, supra, appen. A, p. 166.) Using the pure play method, the assessor identifies publicly traded companies that compete in the same line of business as the property being valued, estimates “the typical capital structure (proportions of debt and equity)” for companies in that line of business, estimates the appropriate cost of debt and equity, and calculates the WACC from this information. (Ibid.)
The assessor attempted to use the pure play method to calculate the WACC for plaintiffs’ properties. John Matthews, a certified property tax appraiser testifying on behalf of the county, explained that he identified eight publicly traded companies in the business of generating electricity, and based his calculations of the WACC on their data. “The best candidates for comparable companies are nonintegrated, single-product companies in a line of business closely related to the asset or property being appraised.” (SBE Assessors’ Handbook, supra, appen. A, p. 166.) Plaintiffs are stand-alone wind energy companies. The county’s witnesses readily admitted that the eight companies identified in their appraisal were not single-product companies in a closely related line of business, nor were they selected because they were similar to the property being appraised. The companies in the assessor’s study are major *733energy companies that derive income from all types of electricity generation; the witnesses presented by the assessor did not know which of them derived income from wind energy.
The WACC is calculated by taking the estimated cost of debt and the estimated cost of equity and combining them, weighting them according to their proportions in the capital structure. (SBE Assessors’ Handbook, supra, at p. 98 & appen. A, p. 178.) The calculation uses after-corporate-income-tax figures. (Ibid.) Because interest on debt is tax deductible, the estimated cost of debt (i.e., the interest rate on debt) must be converted to an after-tax rate. (Ibid.) “The after-tax cost of debt equals the before-tax interest rate multiplied by the combined federal and state marginal corporate income tax rate (i.e., the tax rate paid on each incremental dollar of income).” (Id., appen. A, p. 178.) In his calculations, Matthews determined the after-tax cost of debt by multiplying the before-tax interest rate by the maximum combined federal and state marginal corporate income tax rate: 41 percent.4
John Louden, an auditor-appraiser with the assessor’s office, testified that he took the after-tax capitalization rate calculated by Matthews using the WACC method and converted it into a before-tax rate to be applied to the before-tax income stream being valued. In doing so, he did not use the maximum combined federal and state marginal rate; instead, he calculated an “effective tax rate” for each company. Louden began with the same eight companies used by Matthews, which he believed represented potential buyers of the property being appraised. Using data gleaned from the companies’ 10k reports filed with the Securities and Exchange Commission, Louden calculated for each company a “net income adjusted to DCF [discounted cash flow] analysis income”; apparently, he adjusted the income or loss reflected in the companies’ 10k reports to reach a figure approximating the “net return” prescribed by Rule 8, subdivision (c).
Louden arrived at an “effective tax rate” for each company based on the actual income taxes the companies paid. He then excluded some of the companies for various reasons; he took the rates for the three remaining companies and calculated a weighted average, added a property tax component, and used the resulting rate to convert Matthews’s after-tax discount rate *734to a before-tax discount rate. The result was a before-tax discount rate (“present value factor”) of 11.5 percent for 2006 and 14.75 percent for 2007.5
At the May 12, 2009, hearing, plaintiffs presented their calculations of the before- and after-tax discount rates, which followed a process similar to that followed by the assessor, but used an SBE capitalization rate study as the basis for the after-tax discount rate. Plaintiffs noted that their cash flows differed only slightly from the assessor’s and they agreed with most of the components, with the exception of certain items in Sky River’s 2007 appraisal. Plaintiffs argued for before-tax discount rates of 20.44 percent for 2006 and 21.32 percent for 2007. At the end of the May 12, 2009, hearing, the parties stipulated to the assessor’s cash flow figures, but reserved for argument the issue of the proper capitalization rate and the final property values; they excepted from the stipulation the 2007 assessment for Sky River, which the board set for a further hearing on September 15, 2009.
At the September 15, 2009, hearing, the assessor presented the same method of calculating the before-tax discount rate as he had presented at the earlier hearing, and used it to discount Sky River’s anticipated income stream to present value. Plaintiffs challenged some of the income figures in the assessor’s cash flow data, offered expert testimony to show the assessor’s discount and valuation figures were incorrect, and presented their own before-tax discount rate and final valuation figures.
After the hearings were complete, the board issued its findings of fact, in which it accepted the assessor’s discount rates, finding that the assessor’s methodology for conversion of the after-tax discount rate to a' before-tax discount rate, which relied on “expected” or “effective” corporate income tax rates, complied with Rule 8. Plaintiffs filed complaints in superior court, challenging the rulings of the board. The court tried the issues presented by the complaints: whether the board erred by permitting the assessor to use, in the conversion from an after-tax discount rate to a before-tax discount rate, an “effective” income tax rate that was, in reality, an average rate, rather than a marginal rate, and whether the board erroneously permitted the assessor to use incorrect figures for the capacity payments in his revenue figures for Sky River’s 2007 appraisal. The trial court issued a statement of decision, in which it concluded its standard of review of the board’s action was de novo, the methodology used by the assessor and accepted by the board was legally erroneous, and the use of the wrong figures for the capacity payment was arbitrary and improper. The trial court also concluded that remand to the *735board was not necessary; the errors could be corrected by recalculation, and the properly calculated figures were those presented by plaintiffs in their appraisals at trial.
In this appeal, the county challenges the trial court decision. It contends the proper standard of review was substantial evidence, so the trial court was bound by the findings of the board and could review its decision based only on the administrative record; it was not permitted to take new evidence or make its own factual findings. The county contends the trial court erred in admitting in evidence plaintiffs’ appraisals, which were new evidence, not presented to the board. The appraisals presented by plaintiffs at trial began with the figures from the assessor’s appraisals, but where the assessor used an “effective” tax rate to convert the discount rate from an after-tax rate to a before-tax rate, plaintiffs substituted the maximum marginal combined federal and state income tax rate of 41 percent. The result was a higher before-tax discount rate and a lower property value than those proposed by the assessor. The county contends substantial evidence supported the board’s decision to accept the assessor’s use of an effective or average tax rate and the use of figures for the capacity payment that were based on historical rates superseded by newer contractual rates.
III. Proper Corporate Income Tax Rate to Be Used
In the trial court, plaintiffs did not dispute the county’s cash flow data or its calculation of the after-tax discount rate, with the exception of Sky River’s 2007 assessment, where plaintiffs challenged the figures for income from capacity payments. The primary issue was the income tax rate to be used in converting from an after-tax to a before-tax discount rate. The assessor’s handbook provides that the relevant income tax rate is the expected combined federal and state marginal rate. (SBE Assessors’ Handbook, supra, appen. A, pp. 179-180, fn. 175.) The handbook “is used as a basic guide by tax assessors.” (Exxon Mobil Corp. v. County of Santa Barbara (2001) 92 Cal.App.4th 1347, 1353 [112 Cal.Rptr.2d 751] (Exxon), fn. omitted.) Although assessors’ handbooks are not regulations and do not possess the force of law, they serve as a primary reference and basic guide for assessors, and have been relied upon and accorded great weight in interpreting valuation questions. (Prudential Ins. Co. v. City and County of San Francisco (1987) 191 Cal.App.3d 1142, 1155 [236 Cal.Rptr. 869].) “The interpretations and opinions of an agency administrator, while not controlling upon the courts, constitute a body of experience and informed judgment to which courts and litigants may properly resort for guidance. [Citation.] ‘Because the agency will often be interpreting a statute within its administrative jurisdiction, it may possess special familiarity with satellite legal and regulatory issues. It is this “expertise,” expressed as an interpretation (whether in a regulation or less *736formally . . .), that is the source of the presumptive value of the agency’s views.’ [Citation.]” (Exxon, supra, at p. 1357.)
A marginal income tax rate is the tax rate applicable to the taxpayer’s last dollar of net income. The federal corporate income tax rate increases as taxable income increases (26 U.S.C. § 11), so the rate payable on the highest increment of income is higher than the rate applicable to lower increments; it is also higher than the average rate the taxpayer pays on its entire taxable income. The marginal rate is the rate at which any new income of the taxpayer will be taxed. Consequently, the assessor’s handbook recognizes that, in determining the value of property by using the income approach, the appropriate income tax rate to consider would be the rate at which new income would be taxed, which is the potential purchaser’s combined marginal federal and California income tax rate.6 (SEE Assessors’ Handbook, supra, appen. A, pp. 179-180, fn. 175.)
The California corporate income tax rate is a flat 8.84 percent, subject to a minimum payment (Rev. & Tax. Code, § 23151); thus, this is the marginal California rate. The federal marginal rate to be used in the income approach depends on the taxable income of the potential purchaser. The county did not attempt to calculate the companies’ taxable income for federal tax purposes, from which their federal marginal rates could have been determined.7 Instead, Louden took the income or loss reported by the companies in their publicly available financial statements, deducted capital expenditures, and added depreciation and amortization in order to convert it to a before-tax net return figure as defined in Rule 8, subdivision (c).8 He took the total taxes paid by the company—federal tax, state tax to any state, and foreign tax—and adjusted for tax credits. He then divided the adjusted taxes by his Rule 8, subdivision (c) net return figure, to arrive at his effective income tax rate. Thus, he began with an income figure that may or may not approximate taxable income. Then he lumped together all income taxes paid, to any taxing authority and at any rate, and used the income and tax figures to arrive at some kind of an average rate, which he called the “effective tax rate.” He *737used this rate in lieu of the potential purchaser’s combined federal and California marginal rate in converting the discount rate from an after-tax rate to a before-tax rate.
This average rate did not represent a relevant rate a potential purchaser of the property would take into consideration in determining the price it was willing to pay for the property. It did not represent an estimate of the marginal rate at which the income stream from the subject property would be taxed, since an average rate will always be lower than the marginal rate. Consequently, the county’s attempt to use an average of the potential buyers’ tax rates as an “expected” or “effective” tax rate, in lieu of using the potential buyers’ marginal tax rates was arbitrary and not consistent with the method set out in the assessor’s handbook.
In the administrative proceedings, the board rejected use of the maximum marginal rate, based on a footnote in the assessor’s handbook, which states: “Using the top statutory combined federal and state marginal income tax rate may incorrectly estimate the expected marginal combined rate, which is the relevant rate for the analysis. Because of various tax factors, the expected combined marginal rate may differ from the top statutory combined marginal rate.” (SBE Assessors’ Handbook, supra, appen. A, p. 179, fn. 175.) The board chose to apply the assessor’s average or “expected” tax rate rather than the “maximum tax rate” in the conversion, omitting from its discussion any mention of the “expected marginal combined rate,” which the handbook indicated was “the relevant rate for the analysis.” We do not interpret the handbook’s footnote as a basis for rejecting use of the marginal rate. Rather, we interpret it to mean that the maximum marginal rate may not be the correct marginal rate to use, if the typical potential buyer would not have a taxable income that would be taxed at the highest marginal rate. In that event, the potential buyer’s “expected marginal combined rate,” that is, the marginal rate applicable to the last increment of that potential buyer’s taxable income, would be the appropriate rate to use in the conversion. (Ibid.) It is the potential purchaser’s marginal rate, not some average rate, that is prescribed by the handbook.
We agree with the trial court that the methodology the assessor used and the board approved was legally incorrect. Pursuant to the assessor’s handbook, the appropriate tax rate for the conversion from an after-tax to a before-tax discount rate is the typical potential purchaser’s expected combined California and federal marginal income tax rate. The tax assessor and the board erroneously used an average rate that included taxes paid to other states; the trial court correctly concluded the county’s valuation method was arbitrary and inconsistent with the standards set out in the assessor’s handbook. The trial court did not err in rejecting the county’s methodology.
*738IV. Additional Evidence
Before the board, plaintiffs presented their own calculations to support the discount rate they contended should apply to reach present value of the property under the income approach. In the trial court, however, plaintiffs abandoned their calculations of the cash flow (revenue and expenses) and the after-tax discount rate and accepted the county’s, because their own were not significantly different from the calculations of the assessor. The issue in the trial court was the income tax rate to use in converting from an after-tax discount rate to a before-tax rate. In their presentation to the trial court, plaintiffs submitted appraisals in which they set out the county’s cash flow figures and after-tax discount rate, applied plaintiffs’ proposed conversion rate to the assessor’s after-tax discount rate, then used the resulting before-tax discount rate to calculate the present value of the property. The county contends the trial court’s review of the board’s decision was limited to the administrative record, the trial court was not permitted to admit additional evidence, and plaintiffs’ new appraisals constituted additional evidence that should not have been admitted.
In this matter, plaintiffs are challenging the validity of the method used by the assessor to arrive at the present value of the property of the wind farms. In the trial court, when the taxpayer challenges the method of valuation used by the tax assessor, the issue is a legal one and the trial court “is not restricted to the transcript of the proceedings before the board, but may receive additional evidence bearing on the legal question.” (Norby Lumber Co. v. County of Madera (1988) 202 Cal.App.3d 1352, 1363 [249 Cal.Rptr. 646] (Norby).) The trial court “may take evidence on the validity of the method, and may overturn the assessment if the challenged method of valuation is arbitrary, in excess of discretion, or in violation of standards prescribed by law.” (Dominguez Energy v. County of Los Angeles (1997) 56 Cal.App.4th 839, 852 [65 Cal.Rptr.2d 766].) Consequently, the appraisals submitted by plaintiffs were admissible in the trial court to show that the assessor’s method of calculating the discount rate and the resulting property value was arbitrary, in excess of discretion, or in violation of the applicable legal standards. The appraisals demonstrated the method of converting the after-tax discount rate to a before-tax discount rate that plaintiffs contended was legally correct. Plaintiffs used them both to show that their method was legally correct and to show that the result using the county’s method was a significantly different property value. To the extent they were admitted to demonstrate the invalidity of the assessor’s methodology, we find no error in the trial court’s admission into evidence of plaintiffs’ appraisals.
*739V. Remand
The county contends that, if the trial court was correct in rejecting the board’s decision, this matter must be remanded to the board for a redetermination of the property values and the tax. The trial court concluded the matter did not need to be remanded to the board for further action because the errors could be corrected by recalculation, and the correct calculations of the property values and the property taxes could be found in the numbers provided by plaintiffs’ appraisals. The judgment set out the values to be enrolled for plaintiffs’ property for 2006 and 2007, and ordered the county to calculate and refund the excess tax payments to plaintiffs.
“[W]hen reviewing an equalization determination properly before it in a refund action, a court may correct an assessment and grant a tax refund if value is calculable as a matter of law without remanding to the county board of equalization. [Citations.] [f] However, where a judgment must still be exercised as to value, a remand to the local board of equalization is required. [Citations.]” (Plaza Hollister Ltd. Partnership v. County of San Benito (1999) 72 Cal.App.4th 1, 24 [84 Cal.Rptr.2d 715], italics added, fa. omitted.) “If the board has used an improper method of value or has failed to use proper criteria in valuing the property and there is no evidence or there is a conflict in the evidence from which a proper value can or should be made, the trial court must remand the matter to the board for further proceedings.” (Norby, supra, 202 Cal.App.3d at p. 1366.) Remand “is generally required if the refund determination depends upon an exercise of valuation functions and does not simply involve mathematical computations. [Citation.] The ‘cases establish that the key question concerning remand is whether there remain factual determinations to be made in establishing market value.’ ” (CAT Partnership v. County of Santa Cruz (1998) 63 Cal.App.4th 1071, 1088 [74 Cal.Rptr.2d 652].)
Plaintiffs contend the property values and taxes are calculable from the appraisals presented at trial by plaintiffs, and therefore remand is not required. However, their argument assumes that, as a matter of law, the correct income tax rate to be used in the conversion from an after-tax discount rate to a before-tax discount rate is the maximum combined federal and state marginal income tax rate. That assumption is not correct. The correct rate is the expected combined marginal federal and state income tax rate of a typical potential purchaser of the property in issue. In the administrative proceedings, the county presented evidence of an average income tax rate for the companies it contended were potential purchasers. It presented no evidence of the companies’ expected marginal tax rates. Plaintiffs argued that the maximum marginal rate applied, but presented no evidence that a typical potential purchaser would be taxed at the maximum rate. Plaintiffs argue in this appeal *740that the county’s appraisal documents demonstrate that potential buyers would be in the highest tax bracket. While the county’s appraisals presented some revenue, expense, and tax figures for the eight companies it presented as potential buyers, as plaintiffs themselves recognized in documents filed in the administrative proceeding and the trial court, the county did not present evidence of the companies’ taxable incomes, from which the appropriate marginal rate could be determined. Consequently, there was insufficient evidence to support the assumption that potential purchasers would be taxed at the maximum rate.
Factual issues remain regarding the correct expected combined federal and state marginal income tax rate applicable to a typical potential purchaser of the property in issue. Accordingly, the matter must be remanded to the board so that it can take further evidence and redetermine the appropriate income tax rate to be used in the conversion of the after-tax discount rate to a before-tax discount rate, and thereafter recompute the value to be enrolled and the applicable property tax.
VI. Capacity Payments
The county contends the trial court erred in concluding that substantial evidence did not support the board’s finding that the capacity payment reflected as a revenue item in the assessor’s calculation of the discount rate for the 2007 appraisal of Sky River was an appropriate figure to use in that calculation. A capacity payment is a payment Southern California Edison (Edison), the electric company to which plaintiffs sell the electricity they generate, pays to plaintiffs as an incentive to keep their capacity up at times when the demand is high. The board found the capacity payments reflected in the assessor’s 2007 Sky River appraisal were properly based on historical information known on the appraisal date. The trial court found the assessor “arbitrarily used old historic superseded rates” for the capacity payments reflected for Sky River for 2007, and the rates did not bear any relationship to contract rates in place since 2003, although the correct rates were used for the 2006 appraisal.
On appeal, “ ‘[t]he burden of affirmatively demonstrating error is on .the appellant. This is a general principle of appellate practice as well as an ingredient of the constitutional doctrine of reversible error.’ [Citation.] The order of the lower court is ' “presumed to be correct on appeal, and all intendments and presumptions are indulged in favor of its correctness.” ’ [Citation.]” (State Farm Fire & Casualty Co. v. Pietak (2001) 90 Cal.App.4th 600, 610 [109 Cal.Rptr.2d 256].) “ ‘It is the duty of a party to support the arguments in its briefs by appropriate reference to the record, which includes providing exact page citations.’ [Citations.] If a party fails to support an *741argument with the necessary citations to the record, that portion of the brief may be stricken and the argument deemed to have been waived. [Citation.]” (Duarte v. Chino Community Hospital (1999) 72 Cal.App.4th 849, 856 [85 Cal.Rptr.2d 521].)
Although the county argues in its brief that the trial court erred in finding that the capacity payments reflected in the 2007 Sky River appraisal were not supported by substantial evidence, it fails to cite evidence in the record supporting its assertions. Any reference in an appellate brief to matter in the record must be supported by a citation to the volume and page number of the record where that matter may be found. (Cal. Rules of Court, rule 8.204(a)(1)(C).) This rule applies to matter referenced at any point in the brief, not just in the statement of facts. (Lona v. Citibank, N.A. (2011) 202 Cal.App.4th 89, 96, fn. 2 [134 Cal.Rptr.3d 622].)
At the administrative hearings, there was evidence that the capacity rates are set by a schedule in the contract with Edison; the rates are higher during peak season (June through Sept.) and at certain times of the day. Matthews testified the capacity payments used by the assessor in the 2006 appraisals were calculated “based on the prior three years’ averages for that revenue component.” The county has not pointed to any similar testimony about the 2007 appraisal. There was evidence plaintiffs and Edison entered into a revised capacity agreement in December 2003, and the capacity price was changed for the remainder of the life of the contract. Although this information was provided to the assessor, he used older rates, from the late 1990s and 2000 to 2003, when the capacity revenue was much higher, in his calculations for 2007; he used the current rates in his 2006 calculations. The county has cited no evidence in the record indicating the capacity payment in the 2007 Sky River appraisal was based on current rates or the prior three years’ information; it has cited no evidence that there was any reasonable basis for using superseded data from some earlier period for that appraisal. Consequently, the county failed to meet its burden of demonstrating that the trial court erred in finding that substantial evidence did not support the board’s finding in favor of the county on the capacity payment for the 2007 Sky River appraisal.
DISPOSITION
The judgment of the trial court is reversed, and the trial court is directed to enter a new and different judgment in favor of plaintiffs and against the county, determining that the method used by the board and the tax assessor to calculate the value of the property, and specifically the method used to convert the discount rate from an after-tax rate to a before-tax rate, was arbitrary and invalid. The trial court shall remand the matter to the Assessment Appeals Board for further proceedings to take evidence as needed and *742recalculate the value of the property in accordance with the views expressed in this opinion. The parties shall bear their own costs on appeal.
Cornell, J., and Gomes, J., concurred.
Both Sky River and Mojave are subsidiaries of NextEra Energy, Inc., which is a subsidiary of FPL Group, Inc.
The property in issue includes tangible personal property, improvements and fixtures, but does not include real property, which is taxed separately.
The Legislature has authorized the state’s Board of Equalization (SBE) to prescribe rules and regulations to govern the operation and functioning of local tax assessors and boards of equalization. (Gov. Code, § 15606.) Those regulations are found in the California Code of Regulations, title 18. The parties refer to California Code of Regulations, title 18, section 8, as “State Board of Equalization Property Tax Rule 8” or simply Rule 8. We will follow their lead.
The maximum marginal federal corporate income tax rate is 35 percent. (26 U.S.C. § 11(b).) The flat corporate income tax rate in California is 8.84 percent. (Rev. & Tax. Code, § 23151, subd. (f)(2).) Because state taxes are deductible from income for federal income tax purposes (26 U.S.C. § 164), the combined maximum marginal rate for purposes of calculating the WACC is 41 percent. (SBE Assessors’ Handbook, supra, appen. A, p. 179.)
For Mojave, there was a slightly lower rate for the first few years, due to adjustments for production tax credits for which it was still eligible during those years. Sky River was no longer eligible for those credits.
No evidence was presented or authority cited to indicate income generated by the property in California would be taxed by any other state.
In its reply brief, the county asserts that it used an average tax rate because “there is no publically [iz'c] available data regarding the marginal tax rates of corporate entities.” It does not cite anything in the record in support of this statement, and we have found no evidence to support it.
The discussion of net return in Rule 8, subdivision (c), explains how to calculate the income stream that is to be capitalized when the income approach to value is used. This is a separate calculation from the calculation of the combined federal and state marginal income tax rate to be used in converting the discount rate from an after-tax rate to a before-tax rate. Nothing in Rule 8 suggests that the net return described in subdivision (c) should form the basis for calculating the applicable marginal rate, instead of basing it on taxable income as determined under federal law.