Opinion

RATTIGAN, J. *

Plaintiff and respondent Communications Satellite Corporation (hereinafter Comsat) is a District of Columbia corporation doing *730 business in California. It is also a member of an international consortium which owns and operates commercial satellites orbiting in outer space. As a member of the consortium, Comsat owns an interest in the satellites and receives income from their use. It also derives income from transmitting telecommunications signals to the satellites, and receiving signals from them, at an “earth station” located in California and similar facilities elsewhere.

Defendant and appellant Franchise Tax Board (the Board) assessed certain corporate income taxes against Comsat pursuant to a special formula which resulted in an apportionment of its total income, to California, on the bases of factors which included the value of the satellites and income received by Comsat from their use. Comsat paid the taxes and commenced this action against the Board to recover them, challenging the validity of the special formula. After a nonjury trial on stipulated facts and other evidence, the court found in favor of Comsat and entered a judgment ordering refund of the disputed taxes plus interest. The Board appeals from the judgment.

The Pleadings

Comsat commenced the action by filing a “Complaint For Refund Of Income Taxes” in which it alleged as follows:

Comsat is a corporation “duly organized and existing” pursuant to the Communications Satellite Act of 1962 (47 U.S.C. § 701, et seq.) and the District of Columbia Business Corporation Act. Its principal place of business is located in Washington, D.C. Comsat returned and paid California income taxes to the Board for the calendar years 1970, 1971, and 1972. In May of 1975, the Board notified Comsat that it “proposed to reassess additional income taxes” for these years. Comsat “protested the proposed reassessments,” but the Board “affirmed the proposed assessments in full and denied . . . [Comsat’s] protest.” Comsat paid “the additional amounts” and “additional interest,” and filed timely claims for refunds. The claims were “disallowed” by the Board.

Comsat further alleged: “The income taxes assessed . . . and the interest paid thereon, were illegally, erroneously and arbitrarily determined, assessed and collected . . . , without factual basis for such determination, assessment and collection,” and that they were “void” on specified grounds.

In the prayer of the complaint, Comsat requested “refund” of the additional income taxes paid for the years 1970, 1971, and 1972, “plus all *731 interest paid thereon . . interest “on the said amounts of tax and interest paid . . . from the date of said payments[] to the date of refund”; and general relief.

The Board filed an answer in which it generally denied the allegations made in the complaint.

The Stipulated Facts

After the issues had been joined on the pleadings, the parties filed a written “Stipulation of Facts” executed by their respective attorneys. This document, and the facts stipulated in it, may be summarized and quoted as follows (paragraphing ours):

“Comsat is the United States participant in the International Telecommunications Satellite Consortium (‘Intelsat’), an international joint venture which operates the [szc] global commercial communications satellite system. The joint venture consisted of 83 member countries as of December 31, 1972. Intelsat was established in 1964 by an intergovernmental agreement known as the Agreement ... for a Global Communications Satellite System, which was signed by the United States Government and other governments. Comsat was designated by the President of the United States as the United States signatory to a companion agreement known as the Special Agreement, which governed the operation of Intelsat.”

“The function of the Intelsat system is to relay telecommunications between countries throughout the world. Telecommunications relayed through the . . . system include ‘voice’ (telephone) traffic, ‘record’ traffic of various types (telegraph, telex, facsimile, photogram and data)[,] and television.”

The Intelsat system consists of a “space segment” and a “ground segment.” The assets in the space segment “are owned in undivided shares by members of Intelsat, including Comsat, in varying percentages.” In 1970, 1971, and 1972, Comsat’s “percentage ownership interest” in these assets was “approximately 52.5 percent.” Comsat “serves as manager” of “all facilities” in the space segment.

The assets in the space segment consist of (1) “communications satellites” orbiting in outer space and (2) four “tracking, telemetry[,] and command stations” from which the satellites are “monitored and controlled.” The four tracking stations are located in Maine, Italy, Australia, and Hawaii. The satellites are launched into space from Cape Canaveral, Florida. Each satellite is “located in outer space in a ‘synchronous’ orbit 22,300 miles *732 over a fixed point on the Equator. A ‘synchronous’ orbit is one in which the orbital speed of the satellite is synchronized with the speed of the earth’s rotation so that the satellite appears to remain stationary over a fixed point on earth.. Intelsat satellites are positioned over the Atlantic, Pacific, and Indian Oceans. They never pass over California, even during launch.”

“A communications satellite performs a function similar to an ordinary microwave tower; it receives and transmits signals carrying telecommunications. The principal difference is that microwave towers must be spaced within a few miles of each other because of the curvature of the earth and because the signals travel in a straight line, while a satellite, being 22,300 miles in space, can receive and transmit signals across thousands of miles of the earth’s surface.”

The “ground segment” of the Intelsat system consists of “earth stations in various countries owned by entities in the respective countries. Earth stations have large antennas with supporting electronic equipment. They transmit signals to, and receive signals from, the satellites. . . . Comsat has a 50 percent ownership interest in seven United States earth stations in the Intelsat system.” One of these seven stations (the California earth station) is at Jamesburg, in Monterey County, California, where it was located because of favorable atmospheric and related conditions. The other six earth stations are at Brewster, Washington, and in Maine, West Virginia, Puerto Rico, Hawaii, and Guam. Comsat also owns “a 100 percent interest” in an eighth earth station located at Talkeenta, Alaska. “All of these earth stations became operational before 1970 (with the exception of the Talkeenta station, which did not become operational until July 1970) and remained in operation throughout 1970, 1971[,] and 1972.” Comsat “serves as manager” of all of these earth stations except for the one in Guam.

“In addition to the California earth station, Comsat maintains a small engineering office” in Los Angeles County. Intelsat has no property or employees in California.

“Comsat’s revenues are generated primarily from the leasing of ‘half circuits’ ... to United States international communications common carriers .... A ‘half circuit’ is a two-way communications channel between an Intelsat satellite and an earth station. This channel, when combined with a similar channel (another half circuit) between the satellite and another earth station, permits the establishment of a whole satellite circuit capable of carrying telecommunications between the two earth stations. In order to be able to lease half circuits to the carriers. Comsat leases from Intelsat *733 capacity in the Intelsat satellites and pays ‘use charges’ to Intelsat for such capacity.” All of Intelsat’s revenues are generated by “use charges.”

“Comsat is not involved in transmitting telecommunications between the California earth station [at Jamesburg] and other points in California and does not receive any income from such transmission. Comsat’s activities at the California earth station are concerned only with the leasing of half circuits between the California earth station and Intelsat satellites. The California earth station communicates with Intelsat satellites operating in the Pacific Ocean region.”

“The California earth station operates as follows: (1) A communications signal destined for a foreign country or an offshore United States point is received at the California earth station from its place of origin in the United States through terrestrial facilities (telephone lines, microwave relay systems, etc.) owned and operated by communications carriers other than Com-sat. (2) The earth station transmits the signal to an Intelsat satellite in orbit, which in turn transmits it to an earth station in a foreign country or an offshore United States point. (3) Likewise, communications from foreign countries and offshore United States points are received via satellite by the California earth station and are then transmitted to their point of destination in the United States through terrestrial facilities owned and operated by communications common carriers other than Comsat. The telecommunications transmitted by the California earth station originate at, and the telecommunications received by the California earth station terminate at, locations all over the United States.”

The rates charged by Comsat for half circuits are regulated by the Federal Communications Commission “according to the traditional ‘rate base’ method of regulation. Under this method, Comsat is permitted to make charges designed, but not guaranteed, to generate revenues sufficient to cover its operating expenses (including taxes and depreciation) and provide a fair return on its rate base. The ‘rate base’ consists chiefly of the investment in property devoted to the regulated business.”

In addition to the revenues it derives from leasing half circuits, Comsat receives 52.5 percent of Intelsat’s gross revenues by reason of its (Com-sat’s) 52.5 percent ownership interest in the assets in Intelsat’s space segment.

As of December 31, 1970, Comsat had leased to its customers 2,139 half circuits, of which 405 were between the California earth station and Intelsat satellites (“Califomiahalf circuits”). The “original cost” of Comsat’s interest *734 in the eight U.S. earth stations during 1970 was $43,492,000, of which $3,835,846 (8.8197 percent) was attributable to the Jamesburg station. The “original cost” of Comsat’s interest in the Intelsat satellites in orbit as of December 31, 1970, was $65,893,307.

As of December 31, 1971, Comsat had leased to its customers 2,537 half circuits, of which 468 were California half circuits. The “original cost” of Comsat’s interest in the earth stations during 1971 was $43,810,000, of which $3,933,460 (8.9785 percent) was attributable to the Jamesburg station. The “original cost” of Comsat’s interest in the Intelsat satellites in orbit as of December 31, 1971, was $88,355,540.

As of December 31, 1972, Comsat had leased to its customers 2,971 half circuits, of which 490 were California half circuits. The “original cost” of Comsat’s interest in the eight earth stations during 1972 was $47,477,000, of which $4,114,449 (8.6662 percent) was attributable to the Jamesburg station. The “original cost” of Comsat’s interest in the Intelsat satellites in orbit as of December 31, 1972, was $132,006,758.

“In 1970, 1971[,] and 1972, Comsat derived its income from sources both within and without California, and therefore was subject to the apportionment provisions of section 25101 et seq. of the Revenue and Taxation Code

The Law Involved

As will appear, the issue on this appeal requires identification of the apportionment formula which should be used to determine the corporate income tax owed and payable by Comsat for each of the years in dispute. In the passage last quoted from the stipulation of facts, the parties agreed that the propriety of the formula is controlled by the Uniform Division of Income for Taxation Purposes Act (hereinafter the Uniform Act), which appears in the Revenue and Taxation Code. 1 A discussion of the Uniform Act will be interpolated at this point.

The Uniform Act was proposed in 1957 by the National Conference of Commissioners on Uniform State Laws. It was intended to promote uniformity in allocation practices among the states which impose taxes either on, or measured by, the incomes of corporations. (See Keesling & Warren, *735 California's Uniform Division of Income for Tax Purposes Act (1967) 15 UCLA L.Rev. 156.) The Uniform Act was enacted into California law in 1966. (Stats. 1966, ch. 2, § 7, p. 177.) It has also been adopted by at least 23 other states. (See Table, 62 West's Ann. Rev. & Tax Code (1984 pocket supp.) p. 140.)

The pertinent provisions of the Uniform Act may be summarized as follows: Any taxpayer who receives income from business activity inside and outside the state must “apportion” its net business income to the state. (§ 25121.) The taxpayer’s total net business income is apportioned to California by multiplying the total by a fraction, of which (1) the numerator is the sum of the “property factor,” the “payroll factor,” and the “sales factor,” and (2) the denominator is “three.” (§ 25128.) 2

Each of the three “factors” mentioned is itself a fraction which is defined by a provision in the Uniform Act. The numerator of the property factor fraction is the “average value of the taxpayer’s real and tangible personal property owned or rented and used in this state during the income year.” The denominator is the “average value of [all] the taxpayer’s real and tangible personal property owned or rented and used . . . during the income year.” (§ 25129; see also §§ 25130, 25131.) 3

The numerator of the payroll factor fraction is the “total” compensation paid by the taxpayer in this state during the income year. The denominator is the “total compensation paid everywhere during the income year.” (§ 25132.) 4

The numerator of the sales factor fraction is the sum of “the total sales of the taxpayer” in this state during the income year. The denominator is *736 the sum of the taxpayer’s total sales “everywhere” during the income year. (§ 25134.) 5

The Uniform Act also provides that the taxpayer may request, or the Board may require, variation from the standard formula if it does “not fairly represent the extent of the taxpayer’s business activity” in California; and that the variation may, “if reasonable,” consist of the “exclusion of any one or more” of the three factors described above, or the “inclusion of one or more additional factors which will fairly represent the taxpayer’s business activity in this state,” or the “employment of any other method which will effectuate an equitable allocation and apportionment of the taxpayer’s income” to California. 6

Comsat’s Apportionment Formula

We return to the stipulation of facts. In 1970, 1971, and 1972, pursuant to the provisions of the Uniform Act described above, Comsat “apportioned its net income to California by use of a three-factor formula based on property, payroll, and sales.” We now find it necessary to describe Comsat’s calculations (and later the Board’s) with the complicated details shown in the stipulation. We depart from the parties’ exact language on occasion, but not from the essential facts to which they stipulated.

Comsat’s Calculation of Its Property Factors

In pertinent part, the stipulation continues as follows: “In each of the income years in question, Comsat computed the property factor according to the fraction prescribed by section 25129 . . . .” (See fn. 3, ante.) The numerator of Comsat’s fraction for each year was the value, in the particular year, of Comsat’s interest in real and tangible personal property in Califor *737 nia (i.e., the California earth station at Jamesburg and the office in Los Angeles County). Comsat thus excluded from its numerator (i.e., from the value of its California property) any value representing its interest in the Intelsat satellites. Its denominator was the value, in the particular year, of its interest in all real and tangible personal property everywhere, “including its undivided [52.5%] interest in the Intelsat satellites. . . .” 7

The property factor Comsat calculated from this fraction for 1970, expressed as a percentage, was 2.8296 percent. In fraction format, with translations to the real dollar figures Comsat used for 1970, its calculation of the property factor may be shown as follows:

[[Image here]]

Comsat made similar calculations for each of the other years in question. The property factors it produced were 2.5464 percent for 1971 and 2.2114 percent for 1972.

Comsat’s Calculation of Its Payroll Factors

We here quote the stipulation of facts as follows: “In each of the years in question, Comsat computed the payroll factor according to the fraction prescribed by section 25132 . . . . [See fn. 4, ante.] The numerator consisted of Comsat’s payroll in California and the denominator consisted of its payroll everywhere. . . . The payroll factors thus computed by Comsat, expressed as percentages, were as follows:

[[Image here]]

*738 [[Image here]]

Comsat’s Calculation of Its Sales Factors

Comsat calculated its sales factor for each of the years in question by using a fraction in the format prescribed by section 25134. (See fn. 5, ante.) The numerator of the fraction used in each year was a figure which Comsat called “California earth station revenue” and which excluded any part of the revenue Comsat had received from Intelsat, in that year, by reason of its (Comsat’s) interest in the satellites. Comsat limited its “California earth station revenue” (i.e., its numerator) to a so-called “ground percentage” of the revenue it had received from leasing half circuits at its California earth station in the particular year. It calculated the “ground percentage” by dividing (1) the “net book value” of the real and tangible personal property at the eight earth stations in the year by (2) the “net book value” of all its real and tangible personal property everywhere in the year, including its interest in the satellites. Its calculation of the “ground percentage” in each year may be shown in fraction format as follows (we here quote the stipulation of facts in pertinent part):

[[Image here]]

Comsat next applied the “ground percentage” to the dollar amount of revenue it had received from leasing half circuits at the California earth station in the particular year. It had separately compiled this dollar amount for 1972, but not for 1970 and 1971. It accordingly calculated the amount for 1970 and 1971, and the numerator (California earth station revenue) for each year, by using this two-step formula:

[[Image here]]

*739 [[Image here]]

As previously stated, Comsat had compiled a figure representing the revenue attributable to its leasing of half circuits at the California earth station in 1972. It accordingly calculated its “California earth station revenue” for that year (i.e., the numerator of its sales factor fraction for 1972) by applying its 1972 “ground percentage” to the figure, as follows:

[[Image here]]

According to the stipulation of facts as pertinent here, the denominator of Comsat’s sales factor fraction for each of the three years was its total revenue: i.e., the sum of the revenue it had received “from leasing half circuits” at all the earth stations and its “share of Intelsat revenue” in the particular year. Applying the fraction for each year, Comsat calculated its sales factors as 4.4214 percent for 1970, 3.0098 percent for 1971, and 3.0039 percent for 1972. 8

Comsat’s Apportionment Factor and Further Administrative Proceedings

Having calculated its property, payroll, and sales factors for each of the three years as described, Comsat added the three percentage figures for each *740 year, and divided their sum by three, as prescribed by section 25128. (See fn. 2, ante.) It thereby produced averaged percentages of its “business income . . . apportioned to this state” (see ibid.) which it showed on its corporate income tax returns as 4.1900 percent for 1970, 3.7369 percent for 1971, and 3.4819 percent for 1972. Its California income tax for each year was computed and paid accordingly.

In 1974 and 1975, the Board notified Comsat that it (the Board) “proposed adjustments” for 1970, 1971, and 1972 on the basis that it had “revised” some of the factors Comsat had calculated. The Board’s “revised” factors are described next.

The Board’s Property Factors

In calculating Comsat’s property factor for each of the three years, the Board also used a fraction in the format prescribed by section 25129 (see fn. 3, ante) but with a numerator greater than Comsat’s. The Board enhanced Comsat’s numerator by adding to it (i.e., adding to the value of its California property) a fraction of the value of Comsat’s interest in the satellites in the particular year. The fraction used by the Board for this purpose was calculated as (1) the value (original cost averaged for the year pursuant to sections 25130 and 25131) of Comsat’s interest in real and tangible personal property on the ground in California (i.e., of its interest in the earth station at Jamesburg and the office in Los Angeles County) divided by (2) the value (similarly averaged) of its interest in real and tangible personal property on the ground everywhere (i.e., excluding the value of Comsat’s interest in the satellites).

Application of this fraction to the value of Comsat’s interest in the satellites during the particular year (again, their original cost averaged for the year pursuant to sections 25130 and 25131) produced a result which the Board called “Value of satellites to be apportioned to California.” The numerator of the Board’s property factor fraction was the sum of this value and the value of Comsat’s interest in property on the ground in California (No. 1 above). The Board used the same denominator Comsat had used: i.e., the value, similarly averaged for the year, of Comsat’s interest in all real and tangible personal property everywhere during the year, including the satellites.

Working out this fraction for 1970, the Board calculated a property factor of 4.9187 percent. The calculation may be shown in fraction format, with its elements designated for reference, as follows:

*741 [[Image here]]

With the real dollar figures used by the Board for 1970, its property factor fraction for that year may be shown as follows:

[[Image here]]

Similar calculations by the Board produced property factors of 4.8429 percent for 1971 and 4.8280 percent for 1972.

The Board’s Payroll Factors

The payroll factors Comsat had calculated for 1970, 1971, and 1972 were accepted by the Board without dispute.

The Board’s Sales Factors

In calculating Comsat’s sales factors for 1970, 1971, and 1972, the Board also used a fraction in the format prescribed by section 25134. (See fn. 5, ante.) With some editorial interpolations by this court, the Board’s fraction is partly described in the stipulation of facts as follows:

“For each of the years in question, the Board determined that. . . [Com-sat’s] . . . sales attributable to California (the numerator of the fraction used [by the Board] for the sales factor) consisted of a portion of Comsat’s revenue from Intelsat and a portion of its revenue from . . . [the leasing of half circuits everywhere] .... The denominator consisted of [Comsat’s] total revenue” from the two sources described.

*742 “The formula used by the Board in determining the portion of Comsat’s revenue from Intelsat and the portion of its revenue from . . . [the leasing of half circuits everywhere] ... to be included in the numerator of the sales factor was as follows:

[[Image here]]

“Basing said computation on information then available to it, the Board determined that the sales factor for each of the years in question, expressed as a percentage, was 16.4154 [%].” 9

The stipulation thus describes the Board’s sales factor fraction and the result it yielded, but without illustration. Translating some of the stipulation’s nomenclature for clarity, and designating the elements of the fraction the Board used for the year 1972 only (see fit. 9, ante), we reconstruct the full fraction as follows:

[[Image here]]

*743 The Board’s Apportionment Factor and Further Administrative Proceedings

Adding its property and sales factors to Comsat’s undisputed payroll factor for each year and dividing the sum by three, pursuant to section 25128 (see fn. 2, ante), the Board calculated percentage figures apportioning Com-sat’s income to California at 8.9444 percent for 1970, 8.9709 percent for 1971, and 8.8246 percent for 1972. 10 The Board thereupon notified Comsat that it “proposed to assess additional income taxes” for those years in specified amounts based on the Board’s revised apportionment formula. Comsat “protested the proposed assessments.” The Board denied the protest and gave formal notice of its assessment of the additional income taxes. Comsat paid them and filed claims for refunds, which were also denied. Comsat thereupon commenced this action. The stipulation of facts includes the following language: “Comsat has exhausted its administrative remedies with respect to said refund claims and has timely commenced this action for refund.”

“The Board’s assessment of additional taxes for each of the income years in question was based on its employment of property and sales factors differing from those prescribed [szc] in sections 25129 and 25134, respectively .... In so adjusting the formula used to apportion Comsat’s net income for 1970, 1971 and 1972, the Board relies on the authority granted to it by section 25137 . . . .” (See fn. 6, ante.)

The Trial

Comsat called three witnesses at the nonjury trial. Martin Levine, its “Director of Taxes,” testified that Comsat pays “property taxes in California on its Jamesburg earth station.”

*744 John F. Utley testified as follows: He is associated with an accounting firm which acts as Comsat’s “auditors.” He had been “retained as an expert in this case.” At Comsat’s request, he had prepared a report as to “whether the three factor apportionment formula used by Comsat fairly represented the extent of Comsat’s business activities in California” in the years 1970, 1971 and 1972, and “whether the alternative formula proposed by the Board produced a result which was consistent with Comsat’s business activity in California during the same years.”

Utley further testified that the “calculations” made by Comsat in computing the payroll factor, pursuant to section 25132, were “clearly in accordance with the statute’s requirements. . . .” 11 In the numerator of the property factor fraction for each of the years in question, Comsat “included all its interest in the Jamesburg earth station and the . . . Los Angeles engineering office, but did not include any of the cost of its interest in the Intelsat satellites, which clearly are not property used in any stretch of the imagination in California.” The “property factor as derived by Comsat fairly represents the extent of Comsat’s business activity related to property in California during the years in question.” As to the sales factor, Utley testified that Comsat’s figures “fairly represent the extent of Comsat’s business activity related to sales in California during the years in question,” and that the figures “may be slightly overstated” because “there is no market of Comsat [sic] in California.”

Utley also testified to his opinions that the Board’s calculation of the property factors was “unreasonable”; that it did not “represent the extent of Comsat’s business activity in California”; and that it did not “result in an equitable allocation of income related to property.” He expressed essentially the same opinion as to the sales factors computed by the Board.

A. J. Stotler, the manager of the earth station at Jamesburg, testified that there is “no reason” why Comsat has to have an earth station in California. The signals transmitted and received at the California earth station could be “rerouted” through the earth station at Brewster, Washington.

Subsequent Proceedings

At the conclusion of posttrial arguments by counsel, the trial court stated from the bench that “there will be a finding in favor of . . . Plaintiff [Com- *745 sat] on the complaint ...” The Board requested written findings of fact and conclusions of law. The trial court eventually signed and filed findings and conclusions which we quote in pertinent part as follows:

“The Court finds the following facts to be true ....

“1. The facts as stated in the Stipulation of Facts filed herein are established as true. The Court hereby incorporates the Stipulation of Facts into these Findings of Fact by reference as though fully set forth and finds each fact stipulated to be true.

“5. The property factors computed by Comsat are accurate. It is not true that the standard statutory property factor set forth in § 25129 ... as computed by Comsat does not fairly represent the extent of Comsat’s activity related to property in California, for each of the years in question.

“11. The sales factors computed by Comsat are accurate. It is not true that the standard statutory sales factor set forth in § 25134 . . .as computed by Comsat does not fairly represent the extent of Comsat’s activity related to sales in California, for each of the years in question.

“12. The apportionment factors computed by Comsat are accurate. It is not true that the apportionment factors as computed by Comsat do not fairly represent the extent of Comsat’s business activity in California, in each of the years in question.

“13. The property factor employed by [the] . . . Board ... to apportion Comsat’s income differs from the property factor described in § 25129 . . . in that it includes property not located within the state, namely, a portion of the taxpayer’s interest in communications satellites located in outer space.

“14. The sales factor employed by the Board to apportion Comsat’s income differs from the sales factor described in section 25134 ... in that it includes receipts from the rental or other use of tangible personal property located outside this state, namely, a portion of the receipts from the leasing of Intelsat circuits, and a portion of half circuit rentals attributable to the use of satellites located in outer space.

*746 <6

“Conclusions of Law

“The Court makes the following Conclusions of Law:

“8. The . . . Board bears the burden of proving that the standard statutory provisions do not fairly represent the extent of the taxpayer’s activity in California before it may employ a special allocation and apportionment method. The . . . Board has not carried its burden of proof.

“11. The income taxes determined and assessed to Comsat and collected by the Board in the amounts of $28,307.65 for the year 1970; $136,460.49, for the year 1971; and $151,576.76 for the year 1972; and the interest paid thereon, were illegally, erroneously and arbitrarily determined, assessed and collected by the Board, without factual basis for such determination, assessment and collection, and are void . . . .”

A judgment for the specified amounts was accordingly entered in favor of Comsat. This appeal followed.

Review

The Board contends that this court is not bound by the findings of fact made by the trial court because “this case involves the application of a statute to uncontradicted facts.” This argument must be sustained. The application of a taxing statute to uncontradicted facts is a question of law, and this court is accordingly not bound to accept the trial court’s findings of fact made from the uncontradicted facts shown in the parties’ stipulation and the documentary evidence. (See Southern California Edison Co. v. State Board of Equalization (1972) 7 Cal.3d 652, 659, fn. 8 [102 Cal.Rptr. 766, 498 P.2d 1014]; Anaconda Co. v. Franchise Tax Board (1982) 130 Cal.App.3d 15, 23 [181 Cal.Rptr. 640]; Peninsula Covenant Church v. County of San Mateo (1979) 94 Cal.App.3d 382, 391 [156 Cal.Rptr. 431]; Chase Brass & Copper Co. v. Franchise Tax Bd. (1970) 10 Cal.App.3d 496, 502 [95 *747 Cal.Rptr. 805]; Standard Register Co. v. Franchise Tax Board (1968) 259 Cal.App.2d 125, 129-130 [66 Cal.Rptr. 803].)

Comsat argues that Utley’s testimony supports the trial court’s findings that the results of Comsat’s calculations fairly represented its business activity in California, and that the findings must therefore be sustained on appeal pursuant to the substantial evidence rule. We reject this argument for two reasons.

In the first place, the validity of Comsat’s calculations depends on resolution of the legal question whether the income it received from the use of the satellites was subject to taxation by the State of California. The trial court’s determination of this question is a conclusion of law, not a finding of fact, and it is not binding on this court. (McNeil v. Board of Retirement (1958) 51 Cal.2d 278, 284-285 [332 P.2d 281]; Pac. Pipeline Const. Co. v. State Board of Equal. (1958) 49 Cal.2d 729, 736 [321 P.2d 729].)

Moreover, Utley’s testimony was essentially to the effects that the results of Comsat’-s calculations “fairly represent Comsat’s business activity in California” and that the Board’s results did not. This testimony reflects his views on questions of law based on his interpretation of provisions of the Uniform Act. An expert witness may not properly testify on questions of law or the interpretation of a statute. (See Elder v. Pacific Tel. & Tel. Co. (1977) 66 Cal.App.3d 650, 664 [136 Cal.Rptr. 203]; People v. Clay (1964) 227 Cal.App.2d 87, 98 [38 Cal.Rptr. 431, 100 A.L.R.2d 1421].) Utley’s testimony in these respects was received over a timely objection made on the appropriate ground. (See fn. 11, ante.) The testimony, and the trial court’s findings based on it, are therefore not binding on this court. (See Standard Register Co. v. Franchise Tax Board, supra, 259 Cal.App.2d 125 at pp. 129-130; 6 Witkin, Cal. Procedure (2d ed. 1971) Appeal, § 256, pp. 4247-4248.) We proceed with an independent review of the uncontradicted evidence shown in the stipulated facts and the documents introduced at the trial. (See Southern California Edison Co. v. State Board of Equalization, supra, 7 Cal.3d 652 at p. 659, fn. 8; Peninsula Covenant Church v. County of San Mateo, supra, 94 Cal.App.3d 382 at p. 391.)

The parties expressly stipulated that when the Board recalculated Comsat’s property and sales factors it “relie[d] on the authority granted it by section 25137. ” That statute provides in effect that the Board may deviate from the “allocation and apportionment provisions” of the Uniform Act if their application produces results which do “not fairly represent the extent *748 of the taxpayer’s business activity in this state”; and that the deviation, “if reasonable,” may consist of “(d) The employment of any other method to effectuate an equitable allocation and apportionment of the taxpayer’s income” to California. (See § 25137 as quoted in fn. 6, ante.) In the first instance, therefore, the Board was warranted in recalculating the property and sales factors if the results of their computation by Comsat, pursuant to sections 25129 and 25134, did not “fairly represent” Comsat’s “business activity in this state.”

We have seen that Comsat produced a property factor for each year in the fraction format prescribed by section 25129. The denominator of Com-sat’s property factor fraction was the total value of its real and tangible personal property everywhere, as section 25129 prescribes. (See fn. 3, ante.) In its numerator, however, Comsat did not include any value assigned to the satellites orbiting in outer space. Section 25129 provides in pertinent part that the numerator of the property factor fraction shall include the “value of the taxpayer’s real and tangible personal property owned . . . and used in this state during the income year.” (See ibid.) Comsat’s exclusion of a satellite value from its numerator obviously rested on the premise that the satellites are not located “in this state.”

The pertinent language of section 25129 defining the numerator of the property fraction refers to property “owned” and “used” in this state, not to property “located” here. Moreover, the satellites are not located in California but the Jamesburg earth station is. Because it is, Comsat is conducting “business activity” and generating income in California. There is an invisible, but apparently continuous and very real, connection between the earth station and the satellites. The earth station has a value only because this connection exists, and it is otherwise of no value. Without the connection, the satellites function in outer space to no purpose involving this state. With it, they function in California. The ascription of a “function in California” to the satellites is a recognition of the realities of telecommunications and space technology, not an indulgence in fiction.

Because Comsat owns an interest in the satellites, and because they function in California at and through the Jamesburg earth station, we conclude that they are “tangible personal property owned . . . and used in this state,” by Comsat, within the meaning of section 25129. (See fn. 3, ante.) Com-sat’s exclusion of their value from the numerators of its property factor fractions therefore produced results which, for purposes of the factor, did not “fairly represent the extent of the taxpayer’s business activity in this state” within the meaning of section 25137. (See fn. 6, ante.) This being so, the statute authorized the Board to recalculate Comsat’s property factors *749 by using “any other method” which was “reasonable” and would “effectuate an equitable allocation and apportionment of the taxpayer’s income” to California. (See ibid.)

The Board’s method also involved the use of a fraction format, which was “reasonable” because it was permitted—but not required—by section 25137, subdivision (d). The Board used a numerator which included a value assigned to the satellites. This was done on the theory that the satellites were “tangible personal property owned . . . and used in this state,” as we have just concluded. For purposes of the numerator, the Board fixed their value as a fraction of their total value based on the ratio between (1) the value of Comsat’s real and tangible personal property at the two locations and (2) the total value of its real and tangible personal property everywhere except for the satellites. This step was taken on the theory that the same California-versus-everywhere ratio should be applied in fixing the value of Comsat’s interest in the satellites for purposes of the numerator. The two theories described, and their successive applications in the Board’s numerator, were “reasonable. ” The Board completed its property factor fraction by adding the value of Comsat’s other California property to the numerator and using the total value of its property everywhere as the denominator. Each of these steps was “reasonable” because it was consistent with section 25129. (See fti. 3, ante.) The overall property factor fraction was “reasonable” as the sum of its parts. The Board’s consequent property factor for each year, expressed as a percentage, was accordingly authorized by law and valid in fact.

Comsat also calculated its sales factor for each year in the fraction format prescribed by section 25134 (see fn. 5, ante), but it used a numerator which (1) was limited to a so-called “ground percentage” of the revenue Comsat received from leasing half circuits at the California earth station in the year and which (2) excluded any part of the revenue Comsat received from Intelsat in that year by reason it its (Comsat’s) interest in the satellites. For purposes of the section 25134 numerator, and as pertinent here, the Uniform Act defines the term “sales” as the taxpayer’s “gross receipts” (§ 25120, subd. (e)) and the term “sales of the taxpayer in this state” with reference to the location of the “income-producing activity” in California or its location as between California and “any other state.” (§ 25136.) 12

*750 Comsat apparently calculated its “sales” as “sales in this state,” only, on the theory that the pertinent “income-producing activity” was conducted at the California earth station only; and applied a “ground percentage” to it to exclude any part of the “sales in this state” which were attributable to the satellites (as distinguished from the earth station itself). The latter step involved a subformula which was not expressly authorized by section 25129. Even if use of the subformula was permissible, it had the effect of omitting from Comsat’s “sales in this state” any California half circuit revenue attributable to the use of the satellites. This exclusion, and the parallel omission of any part of the revenue Comsat received from Intelsat by reason of its (Comsat’s) interest in the satellites, turned up a numerator which was substantially less than the sum of its actual “sales in this state” for purposes of 25134.

The denominator of Comsat’s sales factor fraction comported with section 25134 because it was the total of Comsat’s “sales” everywhere i.e., its “gross receipts” from the leasing of half circuits everywhere and its share of satellite revenue received from Intelsat. Because of the diminished numerator, however, the fraction produced a percentage result which, for purposes of the sales factor, did “not fairly represent the extent of the taxpayer’s business activity in this state” within the meaning of section 25137. (See fn. 6, ante.) Again, the Board was authorized by section 25137, subdivision (d), to recalculate the sales factors by employing “any other method” which was “reasonable” and would “effectuate an equitable allocation and apportionment” of Comsat’s income to California. (See ibid.)

The sum, however, was properly the denominator of the sales factor fraction—as both parties used it—because it represented Comsat’s “sales” everywhere. As a numerator reducing it to Comsat’s “sales in this state” only, the Board used a fraction of the sum based on the ratio between the number of half circuits Comsat leased at the California earth station and the number of half circuits it leased everywhere. Stated another way, the Board’s numerator represented a calculation of Comsat’s “sales in this state,” from its “sales” everywhere, in the same proportion its telecommunications traffic at the California earth station bore to its telecommunications traffic everywhere. The calculation was “reasonable” because it fairly apportioned Comsat’s “sales” to California.

The Board’s numerator included part of the revenue Comsat had derived from leasing half circuits. That part was properly characterized as Comsat’s “sales in this state” because the “income-producing activity” which generated it was the leasing of half circuits at the California earth station: i.e., the activity was “performed in this state” within the meaning of section *751 25136, subdivision (a). (See fn. 12, ante.) The balance of the numerator was part of the satellite revenue Comsat received from Intelsat. The “income-producing activity” which generated that revenue was the operation of the satellites themselves. Because they actually “function” at the California earth station (as we have seen), their “income-producing activity” is also “performed in this state” within the meaning of section 25136, subdivision (a). For that reason, the part of Comsat’s “sales” representing revenue from their use was properly characterized as its “sales in this state” within the same meaning.

The calculation of the Board’s numerator was thus “reasonable” because it applied the concept of “sales in this state” which is provided in section 25134 (see fn. 5, ante), and because it did this on a reasonable basis of apportionment. The Board’s use of Comsat’s “sales” everywhere as the denominator was “reasonable” because it was also consistent with section 25134. (See ibid.) As we concluded with regard to the Board’s property factor, its sales factor fraction was “reasonable” as the sum of its parts. The Board’s consequent sales factor, expressed as a percentage, was accordingly authorized by law.

We here reach the fact that the Board calculated the sales factor at 16.4154 percent for each of the three years in question. (See fn. 9, ante.) Comsat complains of this on the appeal, contending in effect that the results are necessarily inaccurate because the figures used in the calculation (Com-sat’s total revenue, the number of half circuits it leased, and the number of half circuits it leased everywhere) differed from year to year.

The parties stipulated that the Board’s calculation of the 16.4154 percent sales factor for each year was “[b]ased on information then available to it . . . .” (See fn. 9 and the accompanying text ante.) This passage is not explained in the stipulation, but it fairly appears that the half circuit numbers “then available” to the Board were for 1972 only; that it computed the 1972 sales factor with those numbers and Comsat’s total revenue for 1972 (see fn. 8); that the result was a 1972 sales factor of 16.4154 percent; and that the Board applied the same factor in 1970 and 1971 because it lacked half circuit numbers for those years. It further appears that the Board’s half circuit numbers for 1972 were wrong, or that it made an arithmetical error in the calculation.

However, all of the pertinent figures for each year are accurately shown in the stipulation of facts or the documentary evidence received at the trial. They demonstrate that their use in the Board’s sales factor fraction would have produced a sales factor of 16.4928 percent for 1972, not 16.4154 *752 percent; and sales factors in excess of 18 percent for each of the other years. Because errors or omissions in the Board’s calculations thus favor Comsat, the latter has no cause to complain. We treat the Board’s sales factors as valid in fact.

The parties exchange opposing contentions as to which of them bears the burden of proof on such issues as the validity of Comsat’s calculation of its property and sales factors. The trial court concluded that the burden was on the Board (see its conclusion of law No. 8, quoted supra), but we need not decide this question. If the Board has the burden, it has been sustained.

The parties argue many other points in their lengthy briefs. The points include such subjects as the effect of the Board’s action in other jurisdictions reached by the Uniform Act and whether the Board has followed its own administrative precedents in this case. Consideration of all of these points is not essential to our disposition of the controversy before us. With two exceptions, we do not reach them.

The first exception is a contention by Comsat that application of the Board’s special apportionment formula denies it (Comsat) due process and equal protection of the laws. This contention was raised in the pleadings, but the trial court did not reach it for obvious reasons. Comsat appropriately makes it in this court by way of defending the judgment. We reject it.

The United States Supreme Court has recently stated the following principles governing state taxation of a business engaged in interstate commerce: “The Due Process Clause of the Fourteenth Amendment imposes two requirements for such state taxation: a ‘minimal connection’ or ‘nexus’ between the interstate activities and the taxing State, and ‘a rational relationship between the income attributed to the State and the intrastate values of the enterprise.’ [Citations.] The tax cannot be ‘out of all appropriate proportion to the business transacted by the . . . [taxpayer] ... in that State. ’ [Citation.]” (Exxon Corp. v. Wisconsin Dept. of Revenue (1980) 447 U.S. 207, 219-220 [65 L.Ed.2d 66, 79, 100 S.Ct. 2109].) Comsat’s maintenance of the California earth station at Jamesburg, and the office in Los Angeles County, satisfies the “nexus” requirement because Comsat has thereby availed itself of the “ ‘substantial privilege of carrying on business’ within the State.” (See id., at p. 220 [65 L.Ed.2d at p. 220].) The second requirement has been satisfied because there is “‘a rational relationship between the income attributed to the State and the intrastate values of the enterprise.’ ” (See ibid.)

*753 The California Supreme Court has pointed out that the state has broad discretion in classifying property for purposes of taxation: “. . . A state tax law is not arbitrary although it ‘discriminate[s] in favor of a certain class ... if the discrimination is founded upon a reasonable distinction, or difference in state policy’ . . . . [t] . . . There exists no ‘iron rule of equality, prohibiting the flexibility and variety that are appropriate’ to schemes of taxation. [Citations.] So long as a system of taxation is supported by a rational basis, and is not palpably arbitrary, it will be upheld despite the absence of ‘a precise, scientific uniformity of taxation.’ [Citations.]” (Amador Valley Joint Union High Sch. Dist. v. State Bd. of Equalization (1978) 22 Cal.3d 208, 234 [149 Cal.Rptr. 239, 583 P.2d 1281].) Comsat is unique by its nature, and its unique character has been fairly taken into account in the special apportionment formula applied by the Board. It has not been shown that the formula denies Comsat due process or equal protection of the laws. (See Board of Medical Quality Assurance v. Superior Court (1980) 114 Cal.App.3d 272, 277 [170 Cal.Rptr. 468].)

We. also consider a claim by Comsat, made for the first time on appeal, that the Board’s application of the Uniform Act amounts to an unconstitutional burden on interstate commerce. This claim fails because it has not been shown that Comsat has been subjected to excessive and duplicative taxation, and that interstate commerce has thereby been discriminated against or otherwise impeded. (See Exxon Corp. v. Wisconsin Dept. of Revenue, supra, 447 U.S. 207 at pp. 227-230 [65 L.Ed.2d 66, 83-85]; Moorman Mfg. Co. v. Bair (1978) 437 U.S. 267, 276-281 [57 L.Ed.2d 197, 206-209, 98 S.Ct. 2340].)

Our independent review of the uncontradicted facts requires reversal of the judgment with directions as ordered below. (See Pacific Motor Transport Co. v. State Bd. of Equalization (1972) 28 Cal.App.3d 230, 243 [104 Cal.Rptr. 558]; 6 Witkin, Procedure, op. cit. supra, Appeal, §§ 546-547, pp. 4487-4489.)

The judgment is reversed. The cause is remanded to the trial court with directions to enter a judgment in favor of defendant and appellant Franchise Tax Board.

Caldecott, P. J., and Poché, J., concurred.

A petition for a rehearing was denied June 29, 1984, and respondent’s petition for a hearing by the Supreme Court was denied August 29, 1984.

Notes

Retired Associate Justice of the Court of Appeal sitting under assignment by the Chairperson of the Judicial Council.

Statutory references are to the Revenue and Taxation Code unless otherwise indicated. The Uniform Act appears as article 2 (commencing with § 25120) of chapter 17 (Allocation of Income, commencing with § 25101) of part 11 (Bank and Corporation Tax Law, commencing with § 23001) of division 2 (Other Taxes, commencing with § 6001).

Section 25128 reads as follows: “All business income shall be apportioned to this state by multiplying the income by a fraction, the numerator of which is the property factor plus the payroll factor plus the sales factor, and the denominator of which is three.”

These sections respectively provide (in full or in pertinent part, as shown):

“25129. The property factor is a fraction, the numerator of which is the average value of the taxpayer’s real and tangible personal property owned or rented and used in this state during the income year and the denominator of which is the average value of all the taxpayer’s real and tangible personal property owned or rented and used during the income year.

“25130. Property owned by the taxpayer is valued at its original cost. . . .

“25131. The average value of property shall be determined by averaging the values at the beginning and ending of the income year. ...”

Section 25132 reads: “The payroll factor is a fraction, the numerator of which is the total amount paid in this state during the income year by the taxpayer for compensation and the denominator of which is the total compensation paid everywhere during the income year.”

Section 25134 provides: “The sales factor is a fraction, the numerator of which is the total sales of the taxpayer in this state during the income year, and the denominator of which is the total sales of the taxpayer everywhere during the income year.”

Section 25137 provides: “If the allocation and apportionment provisions of this act do not fairly represent the extent of the taxpayer’s business activity in this state, the taxpayer may petition for or the Franchise Tax Board may require, in respect to all or any part of the taxpayer’s business activity, if reasonable: [1] (a) Separate accounting; [1] (b) The exclusion of any one or more of the factors; [5] (c) The inclusion of one or more additional factors which will fairly represent the taxpayer’s business activity in this state; or [1] (d) The employment of any other method to effectuate an equitable allocation and apportionment of the taxpayer’s income.”

The stipulation of facts shows that the value Comsat assigned to each item in the fraction was the “original cost” of the respective item. The value Comsat actually used for each item was its “original cost” averaged for the year pursuant to sections 25130 and 25131. (See fn. 3, ante.) This is demonstrated in documentary evidence received at the trial pursuant to a provision in the stipulation.

The step-by-step calculations which produced Comsat’s numerator and denominator for each year, in dollars, do not appear in the stipulation of facts or in the documentary evidence. The numerators and denominators themselves, and the sales factors computed from them by Comsat, are shown in the documentary evidence as follows:

[[Image here]]

The Board calculated an identical sales factor of 16.4154 percent “for each of the years in question” because the “information then available to it” required it to make a one-time calculation with 1972 figures only. The requirement and the result (and errors in the calculation) are described later in this opinion.

The parties’ disparate calculations of the various factors, and their respective results, are tabulated here as follows:

[[Image here]]

This testimony by Utley was received over an objection, by counsel for the Board, that “interpretation” of the statute was properly a function of the court alone. He testified in like manner over a “continuing objection” to the same effect.

As pertinent to the stipulated facts, sections 25120 and 25136 respectively provide:

“25120. As used in Section 25120 to 25139, inclusive, which shall hereafter be referred to as ‘this act,’ unless the context otherwise requires: [fl (e) ‘Sales’ means all gross receipts of the taxpayer . . . .”

“25136. Sales . . . are in this state if: [1] (a) The income-producing activity is performed in this state; or [1] (b) The income-producing activity is performed both in and outside this state and a greater proportion of the income-producing activity is performed in this state than in any other state, based on costs of performance.”