Petro-Canada v. Canada, 2004 DTC 6329, 2004 FCA 158

Date: 20040423

Docket: A-2-03

Citation: 2004 FCA 158

CORAM:        LÉTOURNEAU J.A.

NADON J.A.

SHARLOW J.A.

BETWEEN:

                                                              PETRO-CANADA

                                                                                                                                            Appellant

                                                                           and

                                                    HER MAJESTY THE QUEEN

                                                                                                                                        Respondent

                                          Heard at Ottawa, Ontario on February 9, 2004.

                                  Judgment delivered at Ottawa, Ontario on April 23, 2004.

REASONS FOR JUDGMENT BY:                                                                           SHARLOW J.A.

CONCURRED IN BY:                                                                                        LÉTOURNEAU J.A.

                                                                                                                                      NADON J.A.

Date: 20040423

Docket: A-2-03

Citation: 2004 FCA 158

CORAM:        LÉTOURNEAU J.A.

NADON J.A.

SHARLOW J.A.

BETWEEN:

                                                              PETRO-CANADA

                                                                                                                                            Appellant

                                                                           and

                                                    HER MAJESTY THE QUEEN

                                                                                                                                        Respondent

                                                    REASONS FOR JUDGMENT

SHARLOW J.A.

[1]                In 1992, Petro-Canada claimed a deduction of $46,751,752, which was the amount paid by two joint exploration corporations for certain seismic data and "renounced" to Petro-Canada pursuant to subsection 66(10.1) of the Income Tax Act, R.S.C. 1985 (5^th supp.), C. 1 (as it read before it was repealed in 1997). The Minister reduced the deduction to $8,884,497, which the Minister says was the fair market value of the seismic data at the time of purchase. Petro-Canada appealed to the Tax Court of Canada, without success: Petro-Canada v. Canada, [2003] 2 C.T.C. 2087, 2003 D.T.C. 94 (T.C.C.). Petro-Canada now appeals to this Court.

Facts

[2]                Seismic data is a geophysical record of the results of controlled explosions on or under the ground. It may be used to aid the search for petroleum and natural gas resources. An item of seismic data is called a "line", because the explosions that generate the data are placed along a line on the ground. A line of seismic data is measured in miles or kilometres.

[3]                Seismic data may be obtained by undertaking a program of seismic exploration, a process referred to as "shooting seismic", or it may be purchased from the person who shot it, or from a third party. In the terminology of the industry, the "purchase" of seismic data is actually the purchase of one or more documents, films, tapes or other information storage devices containing seismic data in circumstances that give the purchaser the unlimited right to make exclusive use of the data. Seismic data may also be "purchased" by paying for a copy of the information in some form, coupled with a limited right to make use of the information. The latter transaction is also sometimes referred to as a licence.

[4]                Those who are in the business of exploring for petroleum or natural gas may accumulate large libraries of seismic data as a result of shooting seismic and purchasing seismic data over many years. The value of seismic data may decline somewhat with age. However, old seismic data generally retains some value, in part because of advances in the science of geophysics and the techniques for interpreting seismic data. In addition, there are areas that have been explored seismically in the past, but cannot be explored further in that manner because of urban or industrial development.

[5]                Purchases of seismic data may be facilitated by brokers. Brokers are familiar or may become familiar with seismic data accumulated by exploration corporations. Brokers may also accumulate large libraries of seismic data of their own.

[6]                Generally, the cost of shooting seismic or purchasing seismic data for one's own exploration purposes comes within paragraph (a) of the definition of "Canadian exploration expense" in subsection 66.1(6) of the Income Tax Act, which reads as follows:

[7]                The advantage of having a cost classified as a Canadian exploration expense is that it is added to a pool of expenses ("cumulative Canadian exploration expense"), part or all of which may be deducted immediately or in future years. In certain circumstances, the benefit of the deduction may be transferred to others. Until 1997, a corporation that met the definition of "joint exploration corporation" in the Income Tax Act could transfer joint exploration expenses to a shareholder by means of special election called a "renunciation".

[8]                The term "joint exploration corporation" is defined as follows in subsection 66(15) of the Income Tax Act:

[9]                The term "principal-business corporation" is also defined in subsection 66(15) of the Income Tax Act. The relevant parts of that definition read as follows:

[10]            This appeal concerns renunciations to Petro-Canada by two joint exploration corporations, Phillips P.C. Resources Ltd. (the "Phillips JEC") and Peace Eagle Resources Ltd. (the "CanEagle JEC"). Petro-Canada and Phillips Petroleum Canada Ltd. were the only two shareholders of the Phillips JEC. Petro-Canada and CanEagle Resources Corporation were the only two shareholders of the CanEagle JEC. Petro-Canada, Phillips and CanEagle are all in the business of petroleum and natural gas exploration, development and production.

[11]            In 1991, Phillips had a large library of seismic data, and also a large pool of undeducted Canadian exploration expenses for which it had no immediate use. A plan was devised to finance exploration work for the benefit of both Petro-Canada and Phillips, while at the same time effectively transferring Canadian exploration expenses from Phillips to Petro-Canada.

[12]            The first step in the plan was the creation of the Phillips JEC. The shares of the Phillips JEC were held as to 57% by Phillips and 43% by Petro-Canada. A unanimous shareholders agreement provided that the Phillips JEC would have four directors, two appointed by each shareholder. Directors' resolutions required the approval of all four directors, and there was no provision to break a deadlock. The effect of this corporate structure was that the Phillips JEC was not controlled by either Petro-Canada or Phillips. It was controlled by them jointly.

[13]            The Phillips JEC purchased seismic data from Phillips and began an exploration program under the joint direction of Petro-Canada and Phillips. In 1991, the Phillips JEC purchased approximately 16,227 kilometres of seismic data from Phillips and two of its related corporations for a total purchase price of $26,500,832. In 1992, the Phillips JEC purchased a further 16,152 kilometres of seismic data from the Phillips corporations for $13,251,370.

[14]            The funds required by the Phillips JEC to purchase the seismic data, as well as the funds required for the exploration activities undertaken by the Phillips JEC during the relevant period, were provided by Petro-Canada and Phillips through share subscriptions. The Phillips JEC used the proceeds of the share subscriptions to make loans to the shareholders in proportion to their shareholdings. Those loans were repaid as the Phillips JEC required money. Shareholder loan repayments from Phillips were approximately equal to the cost of the seismic data purchased from the Phillips corporations. Shareholder loan repayments from Petro-Canada were approximately equal to the cost of exploration work undertaken by the Phillips JEC (approximately $20,000,000 in 1991 and approximately $30,000,000 in 1992).

[15]            The Phillips JEC was obliged by contract to renounce to Petro-Canada the first $46,500,000 of Canadian exploration expenses it incurred in 1991, and the first $69,750,000 of Canadian exploration expenses it incurred in 1992. Under the Income Tax Act as it then read, the result of a valid renunciation was that the shareholder would be entitled to claim income tax deductions for Canadian exploration expenses incurred by the joint exploration corporation, as though the expenses had been incurred by the shareholder.

[16]            The creation of the CanEagle JEC was the result of similar arrangements made in 1992 by Petro-Canada and CanEagle. The shares of the CanEagle JEC were held as to 51% by CanEagle and 49% by Petro-Canada. There was a unanimous shareholders agreement which provided that there would be four directors, two appointed by each shareholder. All directors' resolutions were required to be unanimous, and there was no provision to break a deadlock. The CanEagle JEC received $7 million from each of its two shareholders by way of share subscriptions, which it then lent back to the shareholders. Of the shareholder loan repayments, $7 million was used to purchase an undivided one-third interest in 23,748 kilometres of seismic data from CanEagle. The other $7 million was earmarked for exploration activities. The CanEagle JEC was obliged by contract to renounce the first $14 million of Canadian exploration expenses to Petro-Canada.

[17]            For each purchase of seismic data by the Phillips JEC or the CanEagle JEC, there was a contemporaneous opinion of an independent valuator that the agreed purchase price was the fair market value of the seismic data.

[18]            The Phillips JEC and the CanEagle JEC renounced their Canadian exploration expenses to Petro-Canada as they were obliged to do. The renunciations included the cost of the seismic data purchased from Phillips in 1991 and 1992, and the cost of the seismic data purchased from CanEagle in 1992.

[19]            After an audit, the Minister concluded that the fair market value of the seismic data was less than the purchase prices, as illustrated by the following table:

[20]            The Minister reassessed Petro-Canada on the basis that the deduction claimed in respect of the cost of the seismic data could not exceed the fair market value of the seismic data at the time of its acquisition. The result was to reduce the deduction claimed by Petro-Canada from $46,751,752 to $8,884,497.

[21]            Petro-Canada appealed to the Tax Court of Canada. The principal issue raised by Petro-Canada related to its deduction for the renounced expenses.

[22]            The appeal also challenged the Minister's disallowance of certain scientific research and experimental development expenses totalling approximately $700,000. The Crown agrees that Petro-Canada should have been permitted to deduct those scientific research and experimental development expenses. The parties executed a consent to judgment on that point, which was presented to the Judge at the commencement of the trial.

[23]            In the Tax Court, the Crown argued that the cost of the seismic data purchased by the joint exploration corporations did not meet the statutory definition of "Canadian exploration expense". If that argument is correct, the cost of the seismic data to the joint exploration corporations did not qualify for renunciation to Petro-Canada, and Petro-Canada should not have been allowed even the $8,884,497 deduction that was the subject of Petro-Canada's appeal.

[24]            However, the Crown could not and did not argue in the Tax Court that Petro-Canada should be reassessed to disallow the $8,884,497 deduction, because the Crown is not permitted to appeal an assessment: Harris v. Minister of National Revenue, [1965] 2 Ex. C.R. 653, [1964] C.T.C. 562, 64 D.T.C. 5332 (Ex. Ct.), affirmed on other grounds, [1966] S.C.R. 489">[1966] S.C.R. 489, [1966] C.T.C. 226, 66 D.T.C. 5189 (S.C.C.). (The record does not disclose the date of the expiry of the normal reassessment period under subsection 152(4) of the Income Tax Act.)

[25]            The Crown also asserted two alternative arguments to support its position that Petro-Canada could not deduct more than the fair market value of the seismic data (which was how the Minister had in fact reassessed Petro-Canada). One argument was based on paragraph 69(1)(a) of the Income Tax Act, which would apply if, as the Minister assumed, the vendors of the seismic data did not deal at arm's length with the joint exploration corporations. The other argument was based on section 67 of the Income Tax Act, which imposes a test of reasonableness on all deductions in computing income. These provisions are discussed in further detail later in these reasons.

[26]            The Judge dismissed Petro-Canada's appeal. In doing so, he found that only a small portion of the seismic data was acquired for the purpose stated in the definition of "Canadian exploration expense", that the Phillips JEC and the CanEagle JEC did not deal at arm's length with the vendors of the seismic data, and that the fair market value of all of the seismic data was $4,759,464, much less than the Minister's valuation at the time of the reassessment. The Judge also declined to give effect to the consent judgment relating to scientific research and experimental development expenses. Petro-Canada has appealed to this Court.

Issues on appeal

[27]            Petro-Canada argues that the Judge erred in his interpretation of the definition of "Canadian exploration expense", in determining that the joint exploration corporations did not deal at arm's length with the vendors of the seismic data, in determining the fair market value of the seismic data, and in failing to give effect to the consent judgment. I will discuss these issues under the following headings:

A. Interpretation of the statutory purpose test

(1) Background

[28]            In Global Communications Ltd. v. Canada, [1999] 3 C.T.C. 537, 99 D.T.C. 5377 (F.C.A.), this Court held that the cost of seismic data does not meet the definition of "Canadian exploration expense" if the seismic data is purchased for the purpose of selling or licensing it to others. The reasoning behind that conclusion is expressed by Robertson J.A., writing for the Court at paragraph 19 (emphasis added):

[29]            This is an extension by analogy of another decision of this Court, Gulf Canada Ltd. v. Minister of National Revenue, [1992] 1 C.T.C. 183, 92 D.T.C. 6123 (F.C.A.), leave to appeal denied, [1992] S.C.C.A. No. 102 (QL). That case stands as authority for the proposition that lease payments made for land do not meet the definition of "Canadian exploration expense" except to the extent that they permit access to land in a taxation year in which exploration is actually undertaken. The lease payments in issue were paid in respect of approximately four million acres of land, only a small and undetermined portion of which was actually used in any one year for exploration. Hugessen J.A., speaking for the Court, explained as follows (emphasis added):

[30]            In this case, the Judge's approach to the interpretation of "Canadian exploration expense" is based on Global Communications and Gulf Canada. He explains his reasoning as follows at paragraph 30 (emphasis added):

[31]            The question of the actual use of the seismic data purchased by the Phillips JEC and the CanEagle JEC was the subject of many days of evidence. The Judge found as a fact that the Phillips JEC actually used, in its exploration programs, approximately 6.5% of the seismic data purchased from the Phillips corporations, and the CanEagle JEC actually used approximately 4% of the seismic data purchased from CanEagle.

[32]            However, the Judge also recognized that the evidence on actual use was imprecise. For that reason, he allowed a threefold margin for doubt. On that basis, the amount that the joint exploration corporations could have renounced to Petro-Canada (before taking into account the question of fair market value) would be determined as follows:

This calculation suggests that the deduction permitted in respect of the cost of the seismic data should have been $8,591,250, rather than the $8,884,497 allowed by the Minister. (Recall, however, that it is not open to the Crown at this stage to ask that the deduction be reduced, because the Crown cannot appeal an assessment: Harris, cited above.)

(2) Primary purpose

[33]            Petro-Canada argues that the Judge interpreted the statutory purpose test in the definition of "Canadian exploration expense" as though it required exploration to be the "primary" or "dominant" purpose. According to Petro-Canada, this was an error of law by the Judge because a similar approach to the interpretation of the statutory purpose test for interest deductions was rejected in Ludco Enterprises Ltd. v. Canada, [2001] 2 S.C.R. 1082. This is explained in paragraph 50 of the reasons of Iacobucci J., writing for the majority (emphasis added):

[34]            I agree with Petro-Canada that the statutory purpose test in the definition of "Canadian exploration expense" may be met by an expenditure that is made for more than one purpose. However, one of the purposes of the expenditure must be the determination of the existence, location, extent or quality of an accumulation of petroleum or natural gas.

[35]            A corporation like Petro-Canada, the Phillips JEC or the CanEagle JEC, whose business includes exploring for oil and natural gas, might purchase a large block of seismic data merely because some or all of the seismic data might one day be needed or useful for a future exploration program. Petro-Canada argues that there should be a presumption in such cases that the cost of the seismic data qualifies as a Canadian exploration expense. I cannot agree. In my view, such a presumption would not be consistent with the reasoning in Global Communications or Gulf Canada. As I read those cases, the purpose test in the definition of "Canadian exploration expense" requires at least some connection between the purchased seismic data and actual exploration work. Evidence of the actual use of the seismic data for exploration could provide that connection. However, in the absence of such evidence, there must be at least a credible plan for the use of the seismic data in an exploration program within a reasonable time after its acquisition. A hypothetical connection to exploration work that might be done in the future cannot suffice.

[36]            Here, there was evidence of actual use of a small portion of the seismic data for exploration (which the Judge fixed at 6.5% in the case of the Phillips JEC and 4% in the case of the CanEagle JEC), but there was no evidence of any use of the remainder of the seismic data, or a plan to use it. Most of the seismic data was apparently to be retained for an indefinite period of time for possible future reference. A purchase of seismic data merely to create a seismic reference library cannot meet the statutory purpose test.

(3) Tax motive

[37]            Petro-Canada argues that, in applying the statutory definition, the Judge should not have considered Petro-Canada's tax motive in entering into the arrangements relating to the joint exploration corporations, because tax motive is irrelevant. I agree that Petro-Canada's tax motive is not relevant in determining whether there was an exploration purpose for the acquisition of the seismic data by the joint exploration corporations. It does not follow, however, that the Judge's consideration of Petro-Canada's tax motive led to an incorrect conclusion on the question of whether the statutory purpose test was met. Even if the Judge had said nothing about tax motive, he could not reasonably have concluded that all of the seismic data was acquired for the statutory purpose.

(4) Use test

[38]            Referring to the finding of the Judge that only 6.5% of the Phillips JEC seismic data and only 4% of the CanEagle seismic data was actually used for exploration, Petro-Canada argues that the Judge applied a "use" test, rather than the "purpose" test in the statute. According to Petro-Canada, the approach adopted by the Judge required Petro-Canada to prove the actual use of each and every kilometre of each and every line of seismic data purchased. This overstates by a considerable margin the effect of the approach taken by the Judge.

[39]            I have already indicated that I agree with Petro-Canada that an acquisition of seismic data may meet the statutory purpose test even without evidence of actual use of the seismic data for exploration. However, as I read the reasons of the Judge, he did not interpret the statutory purpose as necessarily requiring such proof. Rather, he determined the purpose of the acquisition of the seismic data by following the guidance of Iacobucci J., writing for the Supreme Court of Canada in Symes v. The Queen, [1993] 4 S.C.R. 695, at paragraph 68:

[40]            The Judge determined the purpose of the acquisition of the seismic data by looking at all of the evidence that was offered relating to the nature of the business of Petro-Canada and the joint exploration corporations, the manner in which the exploration business was conducted, the circumstances in which the seismic data was purchased, and its actual use. He noted, among other things, the lack of documentary evidence of any work product by geophysicists engaged by the joint exploration corporations, and the failure of the joint exploration corporations to appoint geophysical staff to review and interpret the seismic data, as contemplated by the unanimous shareholders agreements. He noted that the directors of the joint exploration corporations did not consider what data would be included in the acquisitions. He noted that Petro-Canada's commitment to fund exploration was only for a three year period ending in 1995, and found that the amounts spent for seismic data were not reasonable in relation to the proposed exploration programs of the joint exploration corporations.

[41]            In addition, the Judge noted that, in the case of the seismic data acquired by the Phillips JEC, the selection was done randomly by an employee of Phillips, that no general review of the seismic data was done, and that approximately 40% of the seismic data could not be used in the foreseeable future because it did not relate to land owned by either Petro-Canada or Phillips.

[42]            As for the seismic data acquired by the CanEagle JEC, the Judge noted that it covered a large area of western Alberta and eastern British Columbia, but the CanEagle JEC had plans to explore only in a relatively small area called Evi.

[43]            A review of the Judge's reasons discloses no error in his treatment of evidence of the use of the seismic data to shed light on the purpose of its acquisition.

(5) Interpretation of Gulf Canada and Global Communications

[44]            The final argument for Petro-Canada with respect to the statutory purpose test is that the Judge erred in his reliance on Gulf Canada and Global Communications. It is true that the facts of this case are not the same as the facts of Gulf Canada and Global Communications. However, the Judge concluded that they establish the principles that must guide any interpretation of the purpose test in the definition of "Canadian exploration expense". I agree with him on that point.

(6) Conclusion on the statutory purpose test

[45]            I conclude that the Judge made no error in his interpretation or application of the statutory purpose test in the definition of "Canadian exploration expense", or in finding that only the acquisition of a small part of the seismic data met the statutory purpose test. Therefore, the Minister cannot be compelled to reassess to increase Petro-Canada's 1992 deduction for the cost of the seismic data.

B. Valuation

[46]            I turn now to the Crown's alternative argument that Petro-Canada's deduction should in any event be no greater than the fair market value of the seismic data at the time of acquisition, which the Judge determined to be $4,759,464. If the Crown is correct, then even if the statutory purpose test had been met, Petro-Canada has still been allowed a deduction ($8,884,497) that is greater than its entitlement. I will consider, first, the question of what the fair market value of the seismic data was at the time of acquisition, and then the provisions of the Income Tax Act that, according to the Crown, have the effect of limiting Petro-Canada's deduction to that fair market value.

[47]            Both parties presented in the Tax Court the opinion of an expert as to the fair market value of the seismic data at the time of acquisition. The Judge did not accept the opinion of Petro-Canada's expert because his methodology was not valid. He did not accept the opinion of the Crown's expert because of his reliance on consultants who could not be cross-examined. The Judge made his own analysis of the evidence, and concluded that the fair market value of the seismic data was $4,759,464 in total.

[48]            It is argued for Petro-Canada that the Judge erred in conducting his own valuation of the seismic data. I cannot accept that argument. The fair market value of the seismic data at the time of acquisition is a question of fact. The Judge was free to accept or reject, in whole or in part, any valuation opinions that were offered, and he was free to make his own estimate of value, based on what remained of the acceptable evidence: Connor v. Canada, [1979] C.T.C. 365, 79 D.T.C. 5256 (F.C.A.). His conclusion as to value must stand in the absence of palpable and overriding error: Housen v. Nikolaisen, [2002] 2 S.C.R. 235. The record discloses no such error. Therefore, the remainder of these reasons must be based on the conclusion that the fair market value of the seismic data at the time of acquisition was $4,759,464.

[49]            The Crown's position is that, because the purchase price of the seismic data ($46,751,752) exceeded its fair market value ($4,759,464), the cost of the seismic data to the joint exploration corporations, as determined for income tax purposes, is limited to that fair market value, and Petro-Canada's entitlement to a deduction is similarly limited. The Crown cites paragraph 69(1)(a) of the Income Tax Act, and alternatively section 67 of the Income Tax Act, to support this argument. Those provisions are discussed below.

C. Arm's length

[50]            Paragraph 69(1)(a) of the Income Tax Act applies if, as the Minister assumed and the Judge found, the joint exploration corporations did not deal at arm's length with the vendors of the seismic data (the vendors being, in the case of the Phillips JEC, Phillips or its related corporations, and in the case of the CanEagle JEC, CanEagle). Paragraph 69(1)(a) of the Income Tax Act reads as follows:

[51]            Petro-Canada argues that the Judge misinterpreted the meaning of "arm's length", or misapplied the relevant principles.

[52]            In determining whether, for income tax purposes, parties are considered to deal at arm's length, it is necessary to look first to subsection 251(1) of the Income Tax Act, the relevant parts of which read as follows at the relevant time:

[53]            It is common ground that, for income tax purposes, the joint exploration corporations were not "related" to the vendors of the seismic data. Therefore, the applicable provision is paragraph 251(1)(b).

[54]            There is a large body of jurisprudence dealing with the determination of whether a transaction is between two parties dealing at arm's length. Broadly speaking, the courts have identified three questions that may be used as a framework for analysis; see, for example, Peter Cundill & Associates Ltd v. Minister of National Revenue, [1991] 2 C.T.C. 221, 91 D.T.C. 5543 (F.C.A.). First, is there a common mind directing the bargaining for both parties to the transaction? Second, did the parties to the transaction act in concert without separate interests? Third, did one party to the transaction exercise de facto control over the other?

[55]            The Judge addressed these questions implicitly rather than expressly, and concluded that the joint exploration corporations did not deal with each other at arm's length when entering into the agreement for the purchase and sale of the seismic data. In my view, the evidence justifies that conclusion. The terms of the transactions did not reflect ordinary commercial dealings between vendors and purchasers acting in their own interests. The joint exploration corporations, for example, did not attempt to negotiate a volume discount, as the evidence indicated would be normal for such large acquisitions of seismic data. Neither joint exploration corporation acted independently and in its own interest in entering into the transactions. The terms of the transaction were in fact dictated jointly by Petro-Canada and Phillips (in the case of the Phillips JEC) and jointly by Petro-Canada and CanEagle (in the case of the CanEagle JEC). The joint exploration corporations, for all practical purposes, were indifferent as to the purchase price of the seismic data because, whatever it turned out to be, the shareholders would ensure that the purchase price was funded. Any tax relief relating to the cost of the seismic data would be transferred to Petro-Canada by means of a renunciation.

[56]            In my view, this case cannot be distinguished from [1974] S.C.R. 1144">Swiss Bank Corporation v. Minister of National Revenue, [1974] S.C.R. 1144, [1972] C.T.C. 614, 72 D.T.C. 6470. The question in Swiss Bank was whether interest payments made by a Canadian corporation on loans made by investors in Switzerland were paid at arm's length. The investors were not related to each other. They became involved in the transaction as the result of the promotional activities of [1974] S.C.R. 1144">Swiss Bank Corporation and Swiss Credit Bank. The two Swiss banks each owned 40% of the shares of another Swiss corporation, referred to as S.I.P., which acted as a trustee or agent for the investors. S.I.P. was the sole shareholder of the Canadian corporation to which the investors' funds were lent. Thus, the Canadian borrower was completely captive to the common interests of the lenders, who effectively acted in concert (through S.I.P) in dictating the terms of the loans. Similarly, in this case, the joint exploration corporations were captive to the common interests of their respective shareholders, who acted jointly in dictating the terms upon which the seismic data would be purchased. In my view, the Judge was correct to conclude that the joint exploration corporations did not deal at arm's length with the vendors of the seismic data.

[57]            Petro-Canada also argues that the Judge made three specific errors in his analysis of the arm's length question. The first alleged error is that the Judge paid too much attention to Petro-Canada's tax motive. The transaction was undoubtedly tax motivated, a fact that could not possibly escape the Judge's attention. However, I do not read his reasons as giving that fact any significant weight in reaching the conclusion that the joint exploration corporations did not deal at arm's length with the vendors of the seismic data. There is no merit to this argument.

[58]            The second alleged error is that the Judge focussed on the wrong point in time. Petro-Canada says that the essential terms of the purchase were negotiated at arm's length between Petro-Canada and Phillips, or Petro-Canada and CanEagle, before their respective joint exploration corporations were even formed. According to this argument, the fact that the joint exploration corporations did not participate in the pricing negotiations or attempt to obtain a volume discount is not relevant. I cannot accept this argument. The question asked by paragraph 69(1)(a) is whether the seismic data was purchased by a party dealing at arm's length with the vendor. Necessarily, the relationship between the vendor and the purchaser must be examined at the time of the purchase, not when the terms of the purchase were negotiated.

[59]            The third alleged error is the failure of the Judge to consider what Petro-Canada asserts are "diverging lines of authority" on the question of when parties may be said to be acting in concert without separate interests. This argument is based on the analysis of the arm's length issue in Drouin v. Canada, 2001 D.T.C. 72, [2001] 2 C.T.C. 2315 (T.C.C.), and in particular on the comment in that case that the essential question to be considered in determining whether parties are dealing at arm's length is whether one party controls the other. In my view, this argument is based on a flawed premise. Petro-Canada is assuming that this is a simple case where the shares of a corporation are owned as to 50% each by two parties that are unrelated to each other. If indeed that were the whole story, there would be no basis for concluding that the corporation does not deal at arm's length with either of the two shareholders. That is because, in any normal commercial transaction between the corporation and one of the shareholders, the other shareholder would ensure that the corporation would be able to assert its own interest and would do so. To paraphrase Swiss Bank, there would be some assurance that the terms of any such transaction would reflect ordinary commercial dealing between parties acting in their separate interests. In this case, however, the combined effect of the corporate structure, the unanimous shareholders agreements, the funding arrangements, and the renunciation obligations was that there was no such commercial safeguard.

[60]            For these reasons, I conclude that the Judge made no error in concluding that the joint exploration corporations did not deal at arm's length with the vendors of the seismic data.

D. Section 67

[61]            The Crown argues that, even if the joint exploration corporations dealt at arm's length with the vendors of the seismic data, the fact that the purchase price of the seismic data exceeded its fair market value is sufficient to limit the deduction to that fair market value. The basis of this argument is section 67 of the Income Tax Act, which reads as follows:

This argument was not addressed by the Judge, or by counsel for Petro-Canada in his written or oral argument.

[62]            The leading case on the statutory predecessor to section 67 is Gabco Limited v. Minister of National Revenue, [1968] 2 Ex.C.R. 511, [1968] C.T.C. 313, 68 D.T.C. 5210 (Ex. Ct.). In that case, Cattanach J. stated the following test for the application of this provision:

[63]            Section 67 was considered by this Court in Mohammad v. Canada (C.A.), [1998] 1 F.C. 165, [1997] 3 C.T.C. 321, 97 D.T.C. 5503. The issue was the deductibility of interest paid by a person on a debt used to finance 100% of the purchase price of a rental property. Robertson J.A., writing for the Court, said this at paragraph 28:

[64]            Reasonableness, like value, is a question of fact. In this case, it is a fact upon which the Judge made no finding. While it may be true, as suggested in Mohammad, that paying fair market value for something is prima facie reasonable, I am unable to agree with the Crown that it necessarily follows that paying more than fair market value is unreasonable. There may be circumstances in which a decision to pay more than fair market value for something is a reasonable decision. Considering the test stated in Gabco, I am not persuaded that this is an appropriate case for the application of section 67.

E. Consent Judgment

[65]            The Judge declined to give effect to the consent judgment, which would have increased Petro-Canada's deductions for scientific research and experimental development expenses by approximately $700,000. He reasoned that, because Petro-Canada had been allowed a deduction for renounced Canadian exploration expenses that exceeded its entitlement by much more than $700,000, and the ultimate issue before him was the correctness of the assessment under appeal, he could do nothing more than dismiss the appeal. He could not order the Minister to reassess in accordance with his reasons, because that would have increased the tax payable, which is not a permissible outcome in an income tax appeal (refer to Harris, cited above).

[66]            A consent judgment in respect of an issue in an income tax appeal is not binding on the Tax Court, or this Court. However, it should be given effect unless there is reason to conclude that it is contrary to the Income Tax Act. (One of the objects of Tax Court Rule 170 is to provide an opportunity to explore that possibility.)

[67]            There is no evidence that the consent judgment in this case would have contravened the Income Tax Act in any way. The situation was exactly as if the Judge had decided in Petro-Canada's favour the legal and factual issues implicit in the consent judgment (that is, the issues relating to the deduction for scientific research and experimental development).

[68]            The Judge was correct when he concluded that Petro-Canada had been allowed a deduction that exceeded its entitlement. The only implication of that conclusion was that the Judge could not grant Petro-Canada the remedy it sought, which was an increased deduction for the cost of the seismic data. The Judge was precluded by Harris from requiring the Minister to reduce the deduction because, in effect, that would allow the Crown to appeal the assessment.

[69]            However, the Judge refused to require the Minister to give effect to the consent judgment. Refusing Petro-Canada's rightful claim to the deduction for scientific research and experimental development had the same effect as an order allowing that claim but reducing Petro-Canada's seismic expense deduction by the same amount. It is as though the Judge had allowed, in part, the Crown's appeal of the seismic data deduction. The Judge was doing indirectly what he could not have done directly. In my view, the Judge erred in failing to give effect to the consent judgment.

F. Conclusion

[70]            I would allow this appeal in part, set aside the judgment of the Tax Court of Canada, and replace it with a judgment allowing the appeal in part and referring the reassessment back to the Minister for reassessment on the basis that (1) the deduction for the cost of the seismic data is $8,884,497, and (2) effect must be given to the consent judgment filed in the Tax Court of Canada in respect of scientific research and experimental development expenses. As the appeal succeeded on only a relatively minor issue, the costs of this appeal should be borne by Petro-Canada.

              (s)    "K. Sharlow"                                                                                                                                            J.A.

"I agree

     Gilles Létourneau J.A."

"I agree

     M. Nadon J.A."

FEDERAL COURT OF APPEAL

                            NAMES OF COUNSEL AND SOLICITORS OF RECORD

DOCKET:                                                                               A-2-03

STYLE OF CAUSE:                                                   Petro-Canada v. Her Majesty the Queen          

PLACE OF HEARING:                                                         Ottawa, Ontario

DATE OF HEARING:                                                           February 9, 2004

REASONS FOR JUDGMENT BY:                          Sharlow J.A.

CONCURRED IN BY:                                                          Létourneau J.A.

Nadon J.A.

DATED:                                                                                  April 23, 2004

APPEARANCES:

SOLICITORS OF RECORD:

Deputy Attorney General of Canada

Ottawa, Ontario