Citation: 2011 TCC 232
ALBERTA PRINTED CIRCUITS LTD.,
HER MAJESTY THE QUEEN,
REASONS FOR JUDGMENT
 This is the appeal of Alberta Printed Circuits Ltd. (the “Appellant”) from the reassessments made by the Minister of National Revenue (the “Minister”) under the Income Tax Act (the “Act”) for the Appellant’s 1999, 2000, and 2001 taxation years. By Notice of Reassessment dated September 16, 2005, the Minister reassessed the Appellant for its 1999 taxation year on the basis that the Appellant had overpaid a non-resident corporation, APCI, Inc. (“APCI”), by the amount of $1,066,073, and added that amount to the Appellant’s income pursuant to paragraph 247(2)(a) of the Act. By Notices of Reassessment dated May 18, 2006, the Minister reassessed the Appellant for its 2000 and 2001 taxation years, also on the basis of overpayment to APCI and added the amounts of $1,065,727 and $1,422,775 to the Appellant’s income for those years respectively. The Appellant further appeals the Minister’s assessment pursuant to subsection 247(3) of the Act of penalties totalling $106,572 for 2000 and $142,278 for 2001.
 There are four issues to be decided in this matter:
(1) Whether the reassessments were statute-barred by the application of the limitation period described in Article IX(3) and Article XXVII(3) of the Agreement Between Canada and Barbados for the Avoidance of Double Taxation and the Prevention of Fiscal Evasion With Respect to Taxes on Income and on Capital (the “Treaty”), or whether the limitation period determined by subparagraph 152(4)(b)(iii) of the Act applies;
(2) Whether the Appellant was dealing with ACPI at arm’s length;
(3) Whether the Minister was correct and acting in accordance with paragraph 247(2)(a) of the Act in including the amounts of $894,263, $1,065,727, and $1,422,775, in the Appellant’s income for its 1999, 2000, and 2001 taxation years, respectively; and
(4) Whether the subsection 247(3) penalties were properly assessed.
 It should be noted that, although the reassessed amount for the 1999 taxation year was $1,066,273, the Respondent conceded it should only have been $894,263, effectively granting the Appellant the reduction to match its internal economist’s calculations.
 It was also conceded by both parties that, notwithstanding references to the deductibility of the reassessed adjustments in the pleadings, the issue of the deductibility of such amounts pursuant to paragraph 18(1)(a) of the Act was no longer relevant as the Act’s transfer pricing provisions now work so as to only adjust income.
 The Respondent also conceded that, for the purposes of determining whether the parties were at arm’s length (issue (2) above), the Appellant and APCI, a Barbados corporation, were not de jure controlled by the same group of people and that the only issue in this regard is whether, as a question of fact, they, as persons not related, were dealing at arm’s length or not.
 Finally, it should be noted that there is no dispute between the parties as to the calculation of penalties pursuant to the formula in subsection 247(3) of the Act, if it is in fact determined that there is to be a transfer pricing adjustment pursuant to subsection 247(2) of the Act. The parties accept that such penalties would apply as a consequence of that first provision.
 At the beginning of the trial, the parties submitted a Statement of Agreed Facts, which, together with the evidence submitted during the trial, generally indicates that the relevant facts to consider in this matter are as set out below.
The Canadian Business
 By way of background, the Appellant began its operations in 1984, manufacturing custom prototype circuit boards in Calgary using a then new “serial” manufacturing process developed by Wayne Bamber that allowed the company to provide circuit board designers with samples of their design in very low quantities – with a minimum order being only two pieces. This allowed the Appellant to provide designers with prototype circuit boards at a price cheaper than its competitors, who were then using a different system that required the production of more pieces than designers would need to prove their design, resulting in higher costs. The manufacture of prototypes allowed circuit board designers to ensure their designs were sound before proceeding to mass production. The Appellant’s founders and original owners, Wayne and Geraldine Bamber, were joined by Daniel McMuldroch in the late 1980s.
 Very simply put, the production of custom circuit boards involves the receipt of customer data, its preparation for use in manufacturing (called “set up”), and the manufacture of the board itself. At the time Wayne Bamber developed his “serial” manufacturing process, and until well into the 1980s, the set-up process used in custom circuit board production was entirely manual. When customer data was received a staff member would produce a two-dimensional mock-up for a local photo lab, which would then create a film negative using the mock-up; the negative was then returned to the company and used to manufacture a physical circuit board by applying layers of different products. This process was later automated by the use of computers, with drafting software and electronic manufacturing machinery eliminating the need for photo negatives entirely.
 The Appellant is a Canadian-controlled private corporation. Its fiscal year end is January 31. In 1996, the Appellant’s shares were owned 25 percent by Daniel McMuldroch and 75 percent by Bamber Electronics Ltd. (“BE Ltd.”). BE Ltd.’s shares were owned by Wayne Bamber (60%) and his wife, Geraldine Bamber (40%). In 1996, these three individuals were the directors of the Appellant.
 Mr. McMuldroch first came into contact with the Appellant in 1984 or 1985 when he contracted with the Appellant for the manufacture of circuit boards as a customer. He testified that he worked part-time for the Appellant in 1989 in network administration, web development, and software development. He began to work for the Appellant in full-time employment in 1989, becoming involved with all aspects of the business, including learning the manufacturing processes and working on the development of a bulletin board system to send and receive data to and from clients and of the “set up” system, which, as stated above, was initially a manual one.
 Mr. McMuldroch stated that he first became a 25% shareholder of the Appellant in 1987, despite his evidence that he did not start working for it full‑time until 1989. In 1996, before leaving to reside in Barbados, he continued to own 25% of the Appellant and was involved primarily in software design, web design, and the set-up function.
Set-up and Organization of the Barbados Company
 In 1995, Mr. McMuldroch attended a seminar about doing business in Barbados. He and the Bambers were concerned about asset protection and contemplated moving the manufacturing of the circuit boards outside of Canada due to the high-risk nature of their business.
 Instead, after negotiation between the Bambers and Mr. McMuldroch, it was decided that the “set up” operations ‑ which I will describe in more detail later ‑ would move to Barbados and be carried on by a Barbados corporation.
 Both Mr. Bamber and Mr. McMuldroch acknowledged that they were aware of the tax benefits of operating in Barbados and that their team of accountants and lawyers had considered this when weighing their planning options. However, the evidence of Mr. Bamber was that he was not to be involved as a shareholder, director or officer of the Barbados company and that this venture was something Mr. McMuldroch wanted to do on his own, in hopes of developing the software further and selling it to other customers. Mr. McMuldroch testified that one of the benefits of the move for the Appellant was that it would allow the Appellant to focus on its manufacturing strengths.
 At Mr. McMuldroch’s request, Mrs. Bamber visited Barbados in October of 1996 in order to find a location for the new business and set up a Barbados corporation and bank account.
 Mr. McMuldroch provided her with $10,000 to apply towards the initiation of the company’s incorporation and as a deposit on a lease. The Appellant also provided her with a $1,000 advance for the trip; the evidence is that her hotel, meal, taxi, and film expenses, totalling $1,399 (inclusive of the above-mentioned $1,000 advance), were paid by the Appellant.
 On November 17, 1996, APCI was incorporated in Barbados. APCI is a Barbados International Business Company incorporated under the Barbados Companies Act, and is entitled to a special tax benefit under the Barbados International Business Companies (Exemption From Income Tax) Act (the “IBC Exemption from Income Tax Act”). Its fiscal year-end is December 31.
 Mr. McMuldroch admits the Barbados corporation was incorporated with him as its sole director, president and chief executive officer (“CEO”). Mr. McMuldroch also admitted under cross-examination that, in the application for incorporation, the APCI, Inc. name was the first choice of the parties because the letter APCI were the initials of the “sister” company, which Mr. McMuldroch confirmed meant the Appellant. There is also evidence in Mr. McMuldroch’s application for a work visa in Barbados that he was to be the CEO, that he was then the vice-president of the Appellant (described as a “related company”), that the job required a specialist who was previously employed by the “parent company”, and that he had specialized training received and experience acquired in the employ of APCI’s “affiliated company”. This document contained a declaration sworn by Mr. McMuldroch. He confirmed under cross-examination that references to the parent company, related company, and affiliated company all meant the Appellant. In the declaration, he also confirmed that he was to be supervised by the board of directors of APCI, but that their supervision would be “consultive and infrequent”.
 Mr. McMuldroch obtained loan advances of approximately $140,000 from the Appellant against his expected bonus of approximately $220,000 for 1996. He testified that he repaid it mainly by application of the after-tax bonus amount of $120,381.80 against the loan, as evidenced by the accounting ledger, and the offset of other funds owed to him against it. The evidence is that he used these funds to purchase equipment and goods for APCI, as well as to provide Mrs. Bamber with the $10,000 for use on her above-mentioned trip to Barbados.
 On December 31, 1996, Mr. McMuldroch resigned as a director of the Appellant. BE Ltd. purchased the shares of Mr. McMuldroch in the Appellant on December 31, 1996, for a $500,000 promissory note. This note was repayable over 20 years, with principal instalments of $25,000 per year commencing on December 31, 1997, and ending December 31, 2017, with interest, but which was repaid in full by May 26, 1999, with interest, as per the accounting ledger of the Appellant and as confirmed by Mr. McMuldroch and Mrs. Bamber. There was evidence of the actual payment cheques issued by BE Ltd. to Mr. McMuldroch for those payments and I am satisfied that he was paid in full for those shares.
 Both Mr. and Mrs. Bamber and Mr. McMuldroch testified that they had an agreement to split the profits of the Barbados Company on the basis that two‑thirds would go to the Bambers and one-third would go to Mr. McMuldroch, or to their respective families.
Off-shore Tax Planning Sidebar
 There is strong documentary evidence, confirmed by Mr. McMuldroch and both Gerry Bamber and Wayne Bamber, that the Bambers and Mr. McMuldroch, in consultation with their accountants and lawyers, considered off-shore planning and initially contemplated the creation of a Barbados tax‑exempt corporation. Further, this corporation was to be owned by a Guernsey, Channel Islands, corporation by the name of Ludex, which was actually incorporated on December 2, 1997, and which, in turn, would be owned by two discretionary trusts for which deeds of settlement were executed: one named the Excelon Settlement, the other named the Leduc Settlement, and one holding a one-third interest on behalf of the McMuldroch family, the other a two-thirds interest on behalf of the Bamber family, as confirmed by their oral evidence.
 By a fax dated February 3, 1998, the above-mentioned Canadian accountants wrote to their Channel Islands affiliates and advised that, due to changes in Canadian tax law, the above structure was to be dismantled. Mr. McMuldroch confirmed that this trust approach was abandoned in favour of using “Morville”, as he put it, and Mrs. Bamber confirmed this end to the trust approach as well.
 In the organizational resolutions of APCI which was incorporated on November 17, 1996, Mr. McMuldroch, as the sole director, approved the issuance of 100 fully paid common shares to Morville Ltd. (“Morville”), located in Guernsey in the Channel Islands. Morville was only registered on May 20, 1998. Mr. McMuldroch explained that there were delays in setting up the insurance policies, which are referred to later as the “individual plans”. These plans were to ultimately own Morville’s shares and were the reason a correcting resolution was passed, backdated to November 8, 1996, changing the name of the sole shareholder from Morville to Mr. McMuldroch, as per instructions given by Mr. McMuldroch to the corporate secretary, PriceWaterhouseCoopers, on May 13, 1999, prior to his returning to Calgary. As confirmed by Mr. McMuldroch and the Bambers in their testimony, the need to abandon the initial trust approach to the off-shore planning in February 1998 was the reason for the further delays in finalizing the ownership structure of APCI.
 On May 1, 1998, Mr. McMuldroch transferred his shares in APCI to Morville. Three months later the shareholder’s register of Morville showed that these shares were owned by Nordben Life and Pension Insurance Company Ltd. as follows: 19 shares for Plan A/C-IP00923 and 9 shares for Plan A/C‑IP00919. These shareholdings were confirmed in annual returns filed by Morville in Guernsey in January of 1999 and 2000. The returns demonstrated that the nominees of the two Plans held one share initially and must have transferred it to the Plans, as the latest returns show Plan A/C-IP00923 as having 20 shares and Plan A/C-IP00919 as having 10 shares, all of which is consistent with the 2 to 1 ratio the Bambers and Mr. McMuldroch agreed was the ratio of the profits of APCI they were respectively entitled to. The evidence disclosed that A/C‑IP00919 referred to an insurance plan pursuant to which Mr. McMuldroch had the power to appoint designated beneficiaries, and those he designated were his father, his father’s spouse and their children and remoter issue, as well as his father-in-law and father-in-law’s spouse and their children and remoter issue; in effect, it was an individual plan for the benefit of Mr. McMuldroch’s family and he had control over who the designated beneficiaries would be. Likewise, A/C‑IP00923 referred to an insurance plan that gave Wayne Bamber the right to appoint designated beneficiaries, and those he designated were his father, his father-in-law and their respective spouses and children and remoter issue; it was in effect a plan for the benefit of his family. Mrs. Bamber also confirmed in her testimony that the Plans were for the benefit of the families of Mr. McMuldroch and the Bambers, although she stated that to date the Bambers have received no benefit from their Plan. Mr. Bamber nevertheless acknowledged under cross‑examination that the Plans, as they read, clearly allow the designated beneficiaries to cash in some of the Plan units if they wish; however, he stated that his contact had told him that this was not allowed. In fact, under an endorsement to the Plans, it was in fact their investment manager who had the right to cash in the Plans. Notwithstanding that the terms of the individual plan permit cashing in, Mr. Bamber acknowledged that any investments in the Bamber Plan were for the benefit of his family, which is also clear from the Plans themselves.
 Mr. McMuldroch confirmed on cross-examination that APCI transferred by wire transfer the sum of $1,000,000 from its Barbados account to Morville’s account to be applied to the purchase and creation of the above-mentioned individual plans comprising the life insurance and investment components, two‑thirds for the benefit of the Bamber Plan and one-third for the benefit of his own family Plan. Documentary evidence also showed that each of the Plans had appointed an investment manager, Spread International Investment Limited, and that affiliates of this manager, Spread Services Limited and Co-Sign Services Limited, acted as directors of Morville. Mr. McMuldroch testified that Wayne Bamber had requested confirmation that this money was sent. As mentioned earlier, Mr. McMuldroch testified that it was the delay in funding the Plans and setting them up that caused the confusion in the initial shareholdings of APCI, which started off with Morville even before its incorporation, being shown as the shareholder, but Morville was corrected to read Mr. McMuldroch which was then changed back to Morville in May 1998 after the incorporation of that entity on May 20, 1998, all after the creation of the two Plans, both issued on April 21, 1998.
 It is an agreed fact that between August 13, 1998 and December 22, 2000, APCI paid dividends of $4,283,573 to Morville.
Moving the Set-up Operations to Barbados
 Based on passport stamps, Mr. McMuldroch was in Barbados from December 2 until December 23, 1996, to install computers and software and to ready the facility for operation. Brandi Lewinske, previously employed by the Appellant in its set-up department, arrived December 19 and immediately commenced full-time employment with APCI.
 The only other staff member the Appellant had in its set-up division at the time was Mr. David Wu. Mr. Wu was the newest member of the division. He had less experience and was incapable of performing all of the set-ups required by the Appellant on a daily basis as well as performing his main duties of collecting data from the email attachments sent by clients. As mentioned above, a manual process was used in 1997 before it was replaced by a new, electronic data collection system developed by Mr. McMuldroch.
 After the departure of the set-up division to Barbados, Mr. Wu remained responsible for checking to ensure all data was present in each file before it was transmitted to Barbados. The evidence is that he also checked the set-up, once the completed file was transmitted from Barbados to the Appellant, to ensure there were no obvious errors and to check that the file had not been corrupted during transmission.
 The only other party who performed technology services for the Appellant was Clifford Johnson, who was hired on contract, and later as an employee of the Appellant, to develop a program for regulating the movement of circuit boards through the process line, i.e., the manufacturing process, to maintain the Appellant’s internal network, and to provide technical support to customers experiencing difficulties with the Appellant’s data collection systems.
 Mr. McMuldroch returned to Calgary on December 23, 1996, to take care of outstanding personal issues, including listing his home for sale, but worked evenings and weekends, while in Calgary, completing set-up functions for the Appellant on behalf of APCI before going back to Barbados.
 On January 24, 1997, before departing for Barbados, Mr. McMuldroch caused APCI to purchase any copyrights that Clifford Johnson had in the new printed circuit computer programs that Mr. Johnson had helped develop under Mr. McMuldroch’s direction while under contract with the Appellant. A fee of $500 was paid in order to ensure that Mr. Johnson would not lay claim to any proprietary rights and impede the sale of such software systems in the future. No consideration was paid to the Appellant for any rights it may have had therein.
 In January 1997, Mr. McMuldroch severed his ties with Canada, became a resident of Barbados, and took over the management of APCI. At this time Mr. McMuldroch was the sole director of APCI.
 It is clear from the evidence that, following the move to Barbados, Mr. Bamber was in charge of the manufacturing and chemical processes, Mrs. Bamber was the bookkeeper and in charge of administration, and Mr. McMuldroch focused on software development and data systems as well as the set-up process, that is, the pre-manufacturing stage, and the development of the Appellant’s world website. All parties testified consistently that, after Mr. McMuldroch’s departure to Barbados, there was no one left in Calgary that could perform set-ups for the Appellant, and that they were dependent on APCI during those years for set-ups. In fact, as the Respondent brought to the Court’s attention, the January 31, 1999 financial statements of the Appellant contained an accountant’s note that reads:
7. ECONOMIC DEPENDENCE
The company is dependent upon one of its suppliers for a key component in its production process. Without this supplier, the company would be unable to meet its production schedule.
 Although it is not clear in the context of the note which supplier is referred to, Mr. Bamber confirmed in cross-examination that this was a reference to APCI.
 The Respondent, on cross-examination of Mrs. Bamber, produced a list of employees of the Appellant after the move to Barbados and submitted that they could have also done the set-up work. Mrs. Bamber’s testimony was that they could not have, as they were only capable of doing housekeeping work on data, i.e., inputting and collecting. All of the Appellant’s witnesses were consistent on this point and credible. The Respondent disagrees with them.
 Mr. McMuldroch returned to Barbados on February 2, 1997, and started recruiting staff from the local population. APCI had to purchase as many as ten Lavenir software licenses at one time, Lavenir being the other set-up software used in the set-up operations and which allowed staff to work using an electronically displayed, visual circuit board.
 Recruiting proved difficult in Barbados, particularly when it came to finding employees with sufficient computer skills, and APCI went through a half‑dozen people in the process, selecting potential employees and restarting training every time. The evidence is that it took six months of training before Mr. McMuldroch could have confidence in what the employee was doing, and another six months for the employee to work at the required speed.
Functions Performed by APCI
 APCI and the Appellant entered into four annual contracts setting out APCI’s role and its compensation. The evidence is that these contracts were prepared by Mr. McMuldroch on APCI letterhead and that only two were signed by Wayne Bamber, although Mr. Bamber later testified that he recalled signing all of the contracts. There is no dispute they were used by the parties for determining fees charged and paid. Their interpretation is the source of great dispute between the parties, a matter which I will discuss shortly.
 The contracts were described as being structured in three parts, dealing generally with developing, supplying, and maintaining the database systems software (Part 1), providing set-up functions to replace the Appellant’s set-up department (Part 2), and developing and maintaining the website of the Appellant (Part 3). Mr. McMuldroch stated that the part of the contracts setting out the role of APCI (Part 1) was a “wish-list”, in which the Appellant laid out what it wanted the systems and processes to be in the future.
 The Appellant testified that before the Barbados era, the Lavenir software used was basic and did not allow for multiple versions of circuit boards to be placed on a single panel, the website was not interactive and hence did not allow customers to complete an order form with its required data online, and the different software systems did not communicate with each other, resulting in excess data inputting.
 After the Barbados corporation was founded, APCI received customer data on APCI servers via communications with the Appellant’s server in Calgary by means of the software developed by Mr. McMuldroch. It was through such software that Calgary initially obtained data by electronic means, such as email and, later, an interactive website. The software was developed by Mr. McMuldroch by 1998. At first, it opened customer files, but was further developed to include the automation of set-up steps. Mr. McMuldroch also made improvements to the billing systems and the main database software that interacted with the Lavenir visual computer software improvements that permitted the placement of multiple circuit boards on each panel, instead of using manually produced, layered panels as was the case with the old system. In general, the improvements to the processes reduced data input times through synchronization of information and revolutionized the manufacturing process from a 7 to 10 day timetable, to 24 hours, leading to a radical reduction in the number of panels used per project and large cost savings.
 The set-up process involved the functions of validation, formatting, and panelizing using the Lavenir software, for which APCI staff worked on visual circuit boards using computers.
 Validation, sometimes referred to as verification, involved checking the data received from Calgary to ensure it was properly formatted as imperial or metric and properly sized, which required a check of decimal points. If anomalies were found, APCI would deal directly with the Appellant’s customers’ engineers and a decision would be made as to who would correct these anomalies. Anomalies included diagonal lines appearing on-screen that had to be removed, a lack of drilled holes where such were required, dangling trace lines, i.e., lines not hooked up, or traces improperly joined together.
 Formatting was required for all designs and involved drill dynamics (i.e., setting up drill-hole sizes and assigning the drill sizes used by Calgary to virtual view) and “commonizing” (i.e., ensuring all designs had standard numerical data) as well as the drawing of cutlines for the cutting out of panels.
 Panelizing involved the creation of virtual panels, maximizing space usage, as well as the optimization of the drilling sequence to minimize passes across the panel – i.e., by using an S-curve pattern – for each hole required of a given size.
 The Appellant continued to provide invoicing for both the set-up functions and the actual manufacturing of the circuit boards, shipped the circuit boards directly to its customers, and did not require ACPI to perform any accounting, billing, shipping or collection functions.
Fees paid to APCI
 The Appellant’s witnesses testified that the annual contracts allocated compensation for set-up based on the type of set-up, described as either Protocol 1 (P1) or Protocol 2 (P2), and provided for a square inch fee (or bonus) to compensate for the non-set-up part of the contract, namely Parts 1 and 3 as described above. The Respondent takes the position that the square inch fees were at least partly for the set-up functions performed as well; this is obviously the main dispute between the parties, although the Respondent initially took the position that the square inch fees were also entirely for set-up. The charges were negotiated each year and the contracts show that the bonus/square inch charge was reduced each year through negotiations with Wayne Bamber, due to the decreasing role of APCI after the return of Mr. McMuldroch to Calgary in June 1999 and the fact that APCI did not have personnel with sufficient expertise to fully assist the Appellant’s customers, as well as the fact that the website was substantially completed by late 1998. In addition, Mr. Bamber also testified that the fee was reduced from $0.35 to $0.2285 in the 1998 contract because he simply looked at the number and decided it was too much to pay.
 The Appellant was APCI’s only customer. APCI’s only source of revenue was the amounts paid by the Appellant as its customer and any interest APCI obtained from investing this revenue in term deposit certificates.
 In the annual contracts, the total fees charged for the P1, or Protocol One, panels, which constituted 99% of sales, were stated as follows:*
Contract Term P1 Set-up Fee Fee per Total Square Inch
Square-Inch Fee Paid
Feb 1/97 – Jan 31/98 $60 $0.35 $300,000.00
Feb 1/98 – Jan 31/99 $60 $0.2285 $322,625.13
Feb 1/99 – Jan 31/00 $67 $0.2285 $304,003.35**
Feb 1/00 – Jan 31/01 $67 $0.1143 $293,121.88***
*It should be noted at this time that the manner in which the fees were presented in the annual contracts is a source of dispute between the parties with regard to whether the set-up fees included part of the square inch fees or not. The set-up fee and square inch fee are juxtaposed in the first three contracts and are not in the final contract. This will be discussed later.
**$50,000 was deducted for Mr. McMuldroch’s services that were provided to APCI while he was in the employ of the Appellant, as well as to take into account APCI’s not having staff with sufficient expertise to service the Appellant’s customers. The net figure was reflected as $344,821.92 in the Appellant’s financial statements due to differences in currency conversion rates for reporting purposes.
*** The actual payment of $196,789.79 was calculated on a bumped-up per-square-inch fee of $0.13681948 and was paid in U.S. dollars. The explanation offered at trial was that since APCI was not paid for set‑ups for the month of December 2000, the bump-up was necessary to take into account those services.
 The evidence as to the payment of the base set-up fees during the years in question is substantial and well-informed. Mrs. Bamber, as the bookkeeper for the Appellant, testified in an orderly and very credible manner, substantiating her oral evidence with the volumes of invoice listings in specific ledger accounts, as well as monthly sales reports identifying set-up costs and extras, which included extra drills, mask applications, and mask plots, with all totals summarized below. The following base set-up charges were charged to the Appellant’s Canadian and American customers, who were directly invoiced on a transactional basis, before, during and after the business relationship with APCI:
Fiscal Period Set-Up Fee for P1 Set-up Fee for P2
CDN US CDN US
Feb 1/96 – Jan 31/97 60 48 300 240
Feb 1/97 – Jan 31/98 60 48 300 240
Feb 1/98 – Jan 31/99 60 48* 300 240*
Feb 1/99 – Jan 31/00 67 46 335 230
Feb 1/00 – Jan 31/01 67 46 335 230
Feb 1/01 – Jan 31/02 67 46 335 230
Feb 1/02 – Jan 31/03 67 46 335 230
Feb 1/03 – Jan 31/04 67 46 335 230
Feb 1/04 – Jan 31/05 67** 46 335** 230
*Reduced to $46 for P1 and $230 for P2 due to currency fluctuations
**Reduced to $60 for P1 and $300 for P2 in October 2004 due to currency fluctuations
 It should be noted that while the above rates were consistent with the rates published on the Appellant’s website, as Mrs. Bamber testified, many of the invoices show discounts for customers purchasing a P2 set-up, or discounts as a result of reorders where films could be reused for cost savings in the set-up steps. P2 set-ups only accounted for about 1% of sales.
 With respect to the base set-up fees charged by APCI to the Appellant, the evidence is that there was no mark-up charged by the Appellant to its customers on the fees it paid to APCI. In other words, the fees paid to APCI were the same as those charged by the Appellant to its customers, which were consistent with the level of fees charged by the Appellant to its customers pre‑APCI and post‑APCI. Accordingly, for the 1999, 2000, and 2001 taxation years in dispute, the set-up fees charged by APCI to the Appellant were identical to those charged by the Appellant to its customers in the same period (see the above table, fiscal periods Feb 1/97–Jan 31/98, Feb 1/98–Jan 31/99, and Feb 1/99-Jan 31/00).
 The total amounts paid by the Appellant to APCI during those periods are summarized as follows:
Year Set-Up Fees Paid Square Inch Fees Paid Total Fees Paid*
1999 1,054,572.13 322,625.13 1,377,197.26
2000 1,223,576.30 344,821.92 1,568,398.22
2001 1,124,572.46 293,121.88 1,417,694.34
Total 3,402,720.89 960,568.93 4,363,289.82
*It should be noted that these fees are net of the deductions from the square inch fees taken by the Appellant in 2000, and are adjusted for the average currency exchange rates used by the accountants in the preparation of their financial statements. In other words, these figures are the ones reflected in the financial records and tax returns of the Appellant.
 There is no dispute that the square inch fees were only calculated and paid by the Appellant to APCI at the end of each contract year, which coincided with the Appellant’s fiscal year.
 There is also no dispute that the Appellant paid APCI the total amounts of $1,802,802, $2,046,227, and $2,447,071 in Barbados currency in 1998, 1999, and 2000, respectively, as taken from the Statement of Agreed Facts; however, neither party provided evidence regarding the conversion of Barbados currency numbers into Canadian dollars and, accordingly, the Canadian figures above proven should be used, as the Appellant has proven they were the figures reflected in the preparation of its financial statements and tax returns.
 Nevertheless, those totals clearly demonstrate that revenue of the Appellant increased dramatically during the taxation years in question. The volume of data processed and utilized also tripled during this time.
Return to Canada
 The evidence ‑ from correspondence between Mr. McMuldroch and Wayne Bamber starting May 21, 1998 ‑ is that negotiations ensued following Mr. McMuldroch’s decision to return to Canada, and the Appellant agreed to purchase the software and goodwill (or non-competition) of APCI in order to obtain control of the set-up software and ensure that it was not sold to its competitors. Initial asking prices started at $1 million, with the final purchase price being $50,000, plus $85,000 for goodwill or non-competition, under an agreement between APCI and the Appellant, executed on May 24, 1999, pursuant to which APCI granted the Appellant an option to purchase APCI’s assets. The Appellant exercised this option on December 5, 2000. Although the correspondence suggests that Mr. McMuldroch was negotiating on behalf of Morville, the owner of APCI, he admitted that, as he would ultimately receive 1/3 of the proceeds, he was really just trying to maximize his return and was not negotiating under the authority of Morville but was merely posturing in order to obtain a higher return. It was clear during the negotiations that Mr. Bamber felt he could develop the software himself in Canada, but he wanted control of that version of the software as he had no one to do set-up for him at the end of 1998. There is a serious credibility issue with respect to these so called negotiations, as the evidence clearly establishes that the ultimate beneficiaries of any profit or payouts from APCI were the Bamber and McMuldroch families through the individual Plans’ ownership of Morville. The purchase of APCI’s assets was completed on December 20, 2000.
 Mr. McMuldroch returned to Calgary in 1999 and rejoined the Appellant. The decision was made to wind down APCI by the end of 2000 due to the difficulties and frustration involved in operating in Barbados as a result of the problems with maintaining and training staff and the poorer technological environment generally, as well as due to Mrs. McMuldroch’s health issues.
 Mr. McMuldroch resigned as a director of APCI on June 12, 1999, and on that same date Morville accepted his resignation and appointed Brandi R. Lewinske and itself to be its directors. The evidence is that there was no one in Calgary who could perform the set-up function, other than Mr. McMuldroch, at the time of his return and, accordingly, the set-up contract was continued with APCI in order to allow Mr. McMuldroch time to hire and train Canadian staff in Calgary to take over that function once again. The evidence is that APCI did not perform set-up functions after 2000.
 The evidence was also that Mrs. Bamber caused BE Ltd. to issue a cheque as a deposit on a house being built for Mr. McMuldroch on his return to Calgary, which, she testified, he repaid.
Other Relevant Facts
 During the audit stages, a functional analysis was prepared as a collaboration between the Appellant and APCI (the “Functional Analysis”), with Mr. McMuldroch agreeing that he and Geraldine Bamber were the co-authors. Mr. McMuldroch further admitted he wrote Appendix A to the Functional Analysis, which was a description of the set-up function. Mrs. Bamber confirmed this and stated that Mr. McMuldroch was the one with the set-up knowledge, and thus was obviously more involved with that part of the analysis.
 The evidence, as supported by the financial statements of APCI, also shows that that entity paid tax to the Barbados Government at a rate of 2.5% of its income as a Barbados International Business Corporation. The tax amounts totalled $19,014 for its fiscal year ending December 31, 1997, $31,582 for its fiscal year ending December 31, 1998, $38,195 for its fiscal year ending December 31, 1999, and $53,740 for its fiscal year ending December 31, 2000. There is no evidence as to amounts paid for its 2001 fiscal year. The nature of this “tax” will be discussed later.
 While the ownership of APCI during the taxation years under appeal remains in dispute, the parties agree that BE Ltd. was the sole shareholder of the Appellant and the Bambers were the Appellant’s only officers and directors during that time.
 I propose to deal with the four issues to be determined in this matter in a different order than listed above, due to their interdependence.
Arm’s Length Issue
 In order for the transfer pricing provisions of subsection 247(2) of the Act to apply and for the Minister to be able to assess beyond the normal reassessment period under paragraph 152(4)(b) of the Act, it must first be established that the Appellant and APCI were not dealing at arm’s length. This is evident from the clear wording of the relevant parts of those provisions, as set out below and underlined:
152(4) The Minister may at any time make an assessment, reassessment or additional assessment of tax for a taxation year, interest or penalties, if any, payable under this Part by a taxpayer or notify in writing any person by whom a return of income for a taxation year has been filed that no tax is payable for the year, except that an assessment, reassessment or additional assessment may be made after the taxpayer’s normal reassessment period in respect of the year only if
. . .
(b) the assessment, reassessment or additional assessment is made before the day that is 3 years after the end of the normal reassessment period for the taxpayer in respect of the year and
. . .
(iii) is made as a consequence of a transaction involving the taxpayer and a non-resident person with whom the taxpayer was not dealing at arm’s length . . .
247(2) Where a taxpayer or a partnership and a non-resident person with whom the taxpayer or the partnership, or a member of the partnership, does not deal at arm’s length (or a partnership of which the non-resident person is a member) are participants in a transaction or a series of transactions and
(a) the terms or conditions made or imposed, in respect of the transaction or series, between any of the participants in the transaction or series differ from those that would have been made between persons dealing at arm’s length, or
(b) the transaction or series
(i) would not have been entered into between persons dealing at arm’s length, and
(ii) can reasonably be considered not to have been entered into primarily for bona fide purposes other than to obtain a tax benefit,
any amounts that, but for this section and section 245, would be determined for the purposes of this Act in respect of the taxpayer or the partnership for a taxation year or fiscal period shall be adjusted (in this section referred to as an “adjustment”) to the quantum or nature of the amounts that would have been determined if,
(c) where only paragraph 247(2)(a) applies, the terms and conditions made or imposed, in respect of the transaction or series, between the participants in the transaction or series had been those that would have been made between persons dealing at arm’s length, or
(d) where paragraph 247(2)(b) applies, the transaction or series entered into between the participants had been the transaction or series that would have been entered into between persons dealing at arm’s length, under terms and conditions that would have been made between persons dealing at arm’s length.
 Although there is no definition of the term “arm’s length” in the Act, the statutory provision for determining the existence of arm’s length relationships is found in subsection 251(1) of the Act, the relevant parts of which are reproduced below:
251(1) For the purposes of this Act,
(a) related persons shall be deemed not to deal with each other at arm’s length;
. . .
(c) where paragraph (b) does not apply, it is a question of fact whether persons not related to each other are at a particular time dealing with each other at arm’s length.
 It should be noted that at the beginning of this trial the Respondent conceded that she was only relying on the provisions of paragraph (c) above (although in argument counsel for the Respondent did argue that a case could even be made that the parties are related under paragraph 251(1)(a) as well). There is, then, no need to refer to the provisions of subsections 251(2) to (6), which define and clarify the meaning of “related persons” with respect to paragraph 251(1)(a). In addition, the Appellant referred to section 256 of the Act, which deals with “associated corporations”, and particularly to subsection 256(5.1) of the Act, which reads as follows:
256(5.1) For the purposes of this Act, where the expression “controlled, directly or indirectly in any manner whatever,” is used, a corporation shall be considered to be so controlled by another corporation, person or group of persons (in this subsection referred to as the “controller”) at any time where, at that time, the controller has any direct or indirect influence that, if exercised, would result in control in fact of the corporation, except that, where the corporation and the controller are dealing with each other at arm’s length and the influence is derived from a franchise, licence, lease, distribution, supply or management agreement or other similar agreement or arrangement, the main purpose of which is to govern the relationship between the corporation and the controller regarding the manner in which a business carried on by the corporation is to be conducted, the corporation shall not be considered to be controlled, directly or indirectly in any manner whatever, by the controller by reason only of that agreement or arrangement.
 There was no dispute, however, that section 247 does not contain the words “controlled, directly or indirectly in any manner whatever” and so subsection 256(5.1) is of no relevance here, at least directly; although, as will be evident later, the law does consider both types of control as a factor in determining whether a de facto arm’s length relationship exists between non‑related persons.
 The most recent and leading case dealing with this issue is Canada v. McLarty, wherein the Supreme Court of Canada confirmed the approach taken by the Federal Court in Peter Cundill & Associates Ltd. v. R., (affirmed by the Federal Court of Appeal), which adopted the criteria consistently considered by the courts in order to determine whether or not dealings are at arm’s length ‑ namely, those enunciated in Interpretation Bulletin IT-419. In paragraphs 30 and 31 of that Federal Court decision, Cullen J. said:
30 Whether the parties in this case were dealing at arm’s length is a question to be examined on its own particular facts. Many factors are relevant in the determination of the issue, such as ownership and control of a corporation. However, share control (or absence of it) is not necessarily conclusive; it is only a factor to be considered in determining the question of arm’s length (Robson Leather Co. v. Minister of National Revenue,  C.T.C. 872, 74 D.T.C. 6666, Collier, J. affd  C.T.C. 132, 77 D.T.C. 5106 (F.C.A.)).
31 In Interpretation Bulletin IT-419 Revenue Canada suggested the following factors will determine whether or not dealings are at arm’s length:
(a) the existence of a common mind which directs the bargaining for both parties to a transaction,
(b) parties to a transaction acting in concert without separate interests, and
(c) de facto control.
 In the case at bar, the criterion that best illuminates the nature of the relationship is whether the parties acted in concert without separate interests. Frankly, I find the evidence is overwhelming in favour of finding that the Appellant and APCI were not at arm’s length based on this criterion.
Acting in Concert Without Separate Interests
 The evidence is that Mr. Bamber and Mr. McMuldroch agreed, while they were still directors and shareholders of the Appellant, that a new corporation would be incorporated in Barbados to carry out the set-up functions previously done in-house by the Appellant and to provide other IT and website services for the Appellant, and that the profits of this new corporation were to be divided two-thirds for the benefit of the Bamber family and one-third for the McMuldroch family. An agreement in advance (admitted by all) to divide the profits of APCI is in itself a solid indication that the parties were acting in concert and with the same interest, namely to divide the profits of APCI generated by sales made only to the Appellant. Indeed, an agreement to divide profits between otherwise non-related parties was sufficient to conclude they were not dealing at arm’s length in the case of Fournier (F.) v. Minister of National Revenue, where two non-related parties who were officers, directors, and shareholders of a corporation ‑ one of the two being a minority shareholder having only one vote on a three-man board of directors ‑ agreed to divide the company profits each year by the payment of a dividend. In the case at bar, the principals not only agreed to share the profits, as indicated, throughout the existence of APCI, but also arranged for the profits of APCI to be paid out in the form of a dividend to a holding corporation, Morville, which was owned by an insurance company on behalf of two life insurance policies set up for the benefit of the Bamber and McMuldroch families and reflecting interests of two‑thirds and one-third for those families respectively, and the Bambers and Mr. McMuldroch were themselves part of the group of designated beneficiaries. There is clear evidence Morville used part of those dividends to pay for the insurance policies held by the Plans and totalling $1,000,000.
 Not only was the profit split as agreed, but the structure of this off-shore planning ‑ which replaced earlier off-shore planning involving the use of trusts for the benefit of each of those same parties created in the same tax haven ‑ was clearly agreed to and the agreement was carried out by Mr. Bamber and Mr. McMuldroch acting in concert. They not only sought the same legal and accounting advice together, but directed the implementation of the tax planning jointly as well, corresponding with the same contacts in doing so. Certain items of this planning correspondence appear to have been received individually by them, but correspondence for one references the other.
 In this regard, the parties clearly planned and executed the split of APCI’s profits together. There is, however, a myriad of other evidence that suggests the parties acted in concert without separate interests; their interest being the movement of money off-shore for their mutual benefit. The following are several relevant facts:
1. The evidence is that the Appellant paid a third party, C. Johnson, to develop its software under the direction of Mr. McMuldroch while the latter was a paid employee of the Appellant (before he left for Barbados to join APCI), yet the only party who received any compensation for the software was the third party, who, according to the evidence of Mr. McMuldroch himself, was paid $500 in order to ensure that he relinquished any right he might have had in the copyright. Nothing was paid to the Appellant, which acknowledged paying for the development of the software, but simply took the position that it belonged to Mr. McMuldroch. I do not find this reflective of what arm’s length parties would agree to and must conclude that the only rationale for forgoing compensation for a property right or an interest in property would be that it furthered their common plan.
2. Mrs. Bamber testified that she went to Barbados at the request of Mr. McMuldroch ‑ but also because her husband liked to be in control ‑ in order to set up the incorporation of the Barbados company as well as to find a location for it to operate out of and to enter into a lease for that location. The testimony given was that Mr. McMuldroch gave her $10,000 for such purposes, but, in fact, it was the Appellant who provided the money as an advance to Mr. McMuldroch on account of a year-end bonus for 1996 that was not yet due. In addition, the Appellant expended $1,399 for her hotel accommodation and other costs. In my view, this evidence clearly supports a finding that both parties were involved in ensuring the successful incorporation and set-up of the Barbados corporation (APCI) and hence were working together to further their common interests.
3. The evidence further establishes that the application for incorporation (including APCI’s proposed name), as well as Mr. McMuldroch’s application for a work permit, clearly refers to the Appellant as being one or another of a “related”, “affiliated”, “parent”, or “sister” corporation to APCI. I agree with the Respondent that this speaks to the non-arm’s length relationship between the Appellant and APCI that the parties envisaged.
4. Moreover, the fact that Morville was initially shown as being the sole shareholder of APCI over a year before Morville was even incorporated ‑ a misstep which was later corrected by a resolution dated in June of 1999 changing that shareholder to Mr. McMuldroch ‑ clearly demonstrates that the intention of the parties was to implement the profit-sharing agreement through the Morville entity from the start. The correction only adjusted for the delays in implementing the ultimate structure, as the initial designation of Morville as the sole shareholder would have been impossible.
5. There was also evidence of interchangeable management. Ms. Lewinske testified that as an employee of APCI she always considered herself to be subordinate to both the Bambers and Mr. McMuldroch and she also reported to them. As well, there is evidence that Mr. Bamber was also actively involved in the management of APCI, not only through his involvement in the incorporation and set-up of APCI by sending his wife down to Barbados as mentioned above, but also through his intervention in the loan made to Ms. Lewinske to assist her in purchasing another car in Barbados after Mr. McMuldroch initially declined to do so. In addition, there is documentary evidence that even after he returned to Calgary in June of 1999 to work again for the Appellant as an employee, Mr. McMuldroch was managing Ms. Lewinske and instructing her on how to deal with employment issues. One instance in particular involved an employee complainant, who, obviously also treating Mr. McMuldroch as the boss and the Appellant as head office, wrote to him with complaints regarding Ms. Lewinske’s handling of her matter. This resulted in Mr. McMuldroch writing to Ms Lewinske requesting responses to those complaints. The suggestion by the Appellant’s counsel that such actions are consistent with a party protecting its interests by making sure all was well with its exclusive supplier is not credible in the circumstances. The degree of involvement by Mr. Bamber and Mr. McMuldroch in the management of APCI at times when they allegedly had no legal interest in APCI is not consistent with an arm’s length relationship.
 While I found there was further evidence to support the position of the Respondent that the parties were acting in concert and with the same interests, I believe the above-mentioned evidence provides a more than adequate basis for my conclusions.
 The Appellant’s position on these matters is that the profit-sharing arrangement, the off-shore tax-planning structure, and the fact that the Appellant, as it acknowledged, had no interest in the software purchased by Mr. McMuldroch nor indeed in any matter relating to the relationship were all the result of adversarial and, by implication, arm’s length negotiations between two parties with separate interests, which is further evidenced by the fact that the parties entered into four annual contracts setting out the services to be provided and the parties’ remuneration for those services. Frankly, in my view, the Appellant has simply not rebutted the strong evidence of the Respondent in this regard. The evidence surrounding the annual contracts is that they did not contain many material terms, such as terms relating to liability or warranty with respect to poor work, terms dealing with dispute or jurisdiction issues, or other usual contractual terms one would expect in an international agreement. Moreover, the evidence is that Mr. McMuldroch drafted those contracts, both while he was with APCI and after he had left APCI to return to the Appellant’s employ – a time during which, he admitted, he was no longer an officer, director or shareholder of APCI. His suggestion that he negotiated the last contract on behalf of APCI before he left seems rather lame when the evidence was that he had been negotiating his departure over a year before he left to return to Canada. One would have thought that he would have been putting himself in a conflict of interest by continuing any such negotiations on behalf of an arm’s length party in those circumstances. In my view, such actions speak to the fact that the parties did not consider themselves as being at arm’s length to begin with.
 While I certainly do not doubt that there was much negotiation between Mr. McMuldroch and the Bambers, negotiations leading to the planning, putting into effect, and managing of a common interest do not magically transform those negotiations into evidence of not acting in concert in a common interest or lead to the necessary conclusion that the parties must logically have had separate interests.
 Having determined that the parties acted in concert and without separate interests in creating APCI and carrying on its business and thus that the Appellant and APCI were not dealing at arm’s length during the taxation years in question, it is not necessary to consider the other criteria above mentioned in order to determine whether the parties were at arm’s length or not. However, since the parties went to lengths to discuss the issues, a few brief points are in order.
Common Mind Directing the Bargaining
 In evaluating whether there was a common mind that directed the bargaining for both parties, I note that, while the directing mind was found to be one individual in Cundill, supra, in Swiss Bank Corporation et al. v. Minister of National Revenue, Thurlow J. confirmed that there may be more than one directing mind. At page 5241 of his reasons he stated:
To this I would add that where several parties -- whether natural persons or corporations or a combination of the two -- act in concert, and in the same interest, to direct or dictate the conduct of another, in my opinion the “mind” that directs may be that of the combination as a whole acting in concert or that of any one of them in carrying out particular parts of functions of what the common object involves. . . .
 Many of the facts I have found that substantiate the conclusion that the parties were acting in concert with the same interests in mind also demonstrate that both Mr. McMuldroch and Mr. Bamber were the directing minds of APCI, as they executed a common plan. As I stated earlier, Mr. Bamber had input into both the incorporation and management decisions of APCI, among the latter being those pertaining to car loans for an employee, Ms. Lewinske, who also considered Mr. Bamber to be one of her bosses while she worked for APCI during the taxation years in question. Mr. McMuldroch acted as manager of APCI after he left the Appellant’s employ and the evidence is that he continued to act as a manager after his return, dealing with employee complaints and, in effect, demonstrating that he still supervised Ms. Lewinske and her assistant, Ms. Neil. Both men were the directing minds behind the structuring of the off‑shore plan, including the initial plan involving the use of trusts and the final plan involving the use of insurance policies that will pay out to beneficiaries designated by them and that are effectively controlled by them. These policies own the shares of Morville, the parent company of APCI. In my view, both men were involved in all of the negotiations leading to the creation and execution of a plan for their common interest.
 Importantly, the fact that both of these individuals also carried out some functions separately does not derogate from the fact these functions furthered a common plan and hence the two men can be said to have been the directing minds of the entire plan and its component transactions. Mr. Bamber, who without question had de jure control of the Appellant, stayed in Calgary to continue to run the manufacturing part of the business and supervise the contracting out of the set-up part of the business from the Appellant to APCI, with the assistance of Mrs. Bamber. She continued to ensure that clients were billed for both aspects of the Appellant’s activities and she collected payment after which she then handed over a portion to APCI. While in Barbados, Mr. McMuldroch managed the set-up operation as well as supplied IT and website development services and provided support and maintenance to the Appellant via APCI. Counsel for the Appellant alluded to the fact that these activities were the results of the negotiations between the two sides. Indeed, each individual had separate roles, as well as the roles they shared in the planning and management of the Barbados company, but, when viewed in its totality, the evidence in this case clearly establishes that both Mr. Bamber and Mr. McMuldroch were the controlling minds that directed the entire plan, structure, and division of profits of APCI.
De Facto Control
 The third criterion for determining whether two parties were at arm’s length is whether one party had de facto control over the other.
 The Respondent takes the position that the Appellant had de facto control over APCI either because Mr. Bamber had de facto control over APCI or because Mr. Bamber and Mr. McMuldroch acting jointly had de facto control over APCI.
 There is no dispute between the parties that the Supreme Court of Canada’s decision in Duha Printers (Western) Ltd. v. The Queen, stands for the principle that de jure control requires that the party allegedly in control have the authority to elect the board of directors or have the clear right and ability to effect a significant change in the board of directors. The Appellant takes the position that de facto control, like de jure control, requires the same ability, relying for this on the case of Silicon Graphics Ltd. v. The Queen, and argues there is no evidence that anyone other than Morville or its insurance company parent had this power. The Appellant also suggests that the application of the de facto control tool ‑ as defined in subsection 256(5.1) of the Act ‑ is reserved for only those cases involving associated corporations, as the concept of de facto control is imported into those provisions of the Act where the words “controlled, directly or indirectly in any manner whatever” are used. The Appellant argues that these words are not found in the wording of the transfer pricing provisions of section 247 of the Act.
 In this regard, I can only partially agree with the Appellant.
 Firstly, I do not agree with the Appellant that the application of the de facto control test can be limited to solely those cases involving associated corporations and the application of subsection 256(5.1) of the Act. I acknowledge the Appellant’s reference to Iacobucci J.’s statement in paragraph 52 of the Duha decision:
52 . . . In fact, while it is not directly relevant to the outcome of this appeal, I would observe nonetheless that Parliament has now recognized the distinction between de jure and de facto control, adopting the latter as the new standard for the associated corporation rules by means of s. 256(5.1) of the Income Tax Act . . .
 However, I do not take the learned Justice’s statement to mean that the adoption of the de facto control test as the standard for the associated corporation rules in subsection 256(5.1) excludes its application outside those specific rules. The McLarty case, decided ten years after Duha by the Supreme Court of Canada, recognized de facto control as being one of several criteria for determining whether or not an arm’s length relationship exists. This test has had ample use in that connection by the Federal Court of Appeal and Tax Court of Canada in numerous cases, i.e., in Cundill, earlier referenced, and in Petro‑Canada v. The Queen, which dealt with whether the purchase of certain seismic data was made from an arm’s length party for the purposes of subsection 69(1) of the Act. There is simply no requirement that the words “controlled, directly or indirectly in any manner whatever” must be found in the relevant section for the de facto test to apply. Indeed, it applies anywhere in the Act where a determination of “arm’s length” must be made, except to the extent that its application is qualified by other sections ‑ i.e., such as in the case of associated corporations under subsection 256(5.1) of the Act, which provides an exception to the application of the test in certain commercial arrangements such as franchise agreements.
 Secondly, the test articulated by the Appellant is only partially correct. In fact, in Silicon Graphics, Sexton J.A. stated at paragraph 67:
67 It is therefore my view that in order for there to be a finding of de facto control, a person or group of persons must have the clear right and ability to effect a significant change in the board of directors or the powers of the board of directors or to influence in a very direct way the shareholders who would otherwise have the ability to elect the board of directors.
 Clearly, the test is not as restrictive as counsel for the Appellant argued. If it can be demonstrated that a party has the clear right and ability to effect a change in the powers of the board of directors or to influence in a very direct way the shareholders who would otherwise have the ability to elect the board of directors, then such party has de facto control.
 If one accepts the Appellant’s position that neither the Appellant nor Mr. Bamber had de jure control of APCI, i.e., that they had no clear right or ability to effect a significant change in the board of directors, then it stands to reason that the distribution of profits by the corporation must have been decided either by the board of directors of the corporation or by the shareholders usurping the board’s powers through an agreement such as a unanimous shareholders agreement or otherwise as allowed by the constating documents, of which no evidence exists in this case, or by means of external agreements or arrangements giving persons the clear right and ability to effect a significant change in the powers or composition of the board of directors or to influence in a very direct way the shareholders who would otherwise have the ability to elect the board of directors. Here we have a situation where Mr. Bamber and Mr. McMuldroch, acting in concert, agreed that the profits of APCI would be divided in two-third and one-third proportions for the benefit of their respective families, which include themselves. If this agreement does not have a significant effect on the powers of the board of directors to distribute profits of the corporation or evidence the ability of Mr. Bamber and Mr. McMuldroch to influence the shareholders (technically speaking) of Morville, the parent company of APCI, in a very direct way, I do not know what does. Clearly, Mr. Bamber alone, having a two-thirds interest, can be said to have de facto control and, alternatively, Mr. Bamber and Mr. McMuldroch, acting jointly and with a combined 100% interest, can be said to also have de facto control in the circumstances.
 Moreover, the evidence is that the shares of Morville were owned by Nordben Life and Pension Insurance Company Ltd., with 20 shares attributable to Plan A/C-IP00923 and 10 shares to Plan A/C-IP00919, representing a two‑thirds and one-third split in ownership. As indicated earlier, the evidence is also that the first individual plan, No. IP-00919, was a plan for the benefit of the Bamber family, with respect to which Mr. Bamber controlled the right to designate the beneficiaries, while the second individual plan, No. IP-00923, was for the benefit of Mr. McMuldroch’s family, and with respect to which Mr. McMuldroch had the right to designate the beneficiaries. There was also documentary evidence that for each of these individual plans the same investment manager was appointed by both Mr. Bamber and Mr. McMuldroch. This manager not only reported to them but also had the right to surrender or cash in all or part of the investments under the Plans. In my view, notwithstanding the air of direct ownership of these shares by a foreign insurance company, they were held specifically for the account of the two families under plans created with the involvement of Mr. Bamber and Mr. McMuldroch and by their signatures for the benefit of their families, and the shares were beneficially owned by Mr. Bamber and Mr. McMuldroch and controlled by them. The insurance company was no better than a trustee acting on their behalf and they clearly had beneficial ownership of Morville’s shares. The structure of this off-shore planning at the very least serves to support the fact that Mr. Bamber and Mr. McMuldroch had significant influence over Morville, which, as its sole shareholder, had the right to appoint the directors of APCI. In fact, there is ample evidence that they exercised their significant influence by having Brandi Lewinske appointed as a director of APCI as a local representative on the board of directors. Ms. Lewinske was an employee of APCI and acknowledged she took the appointment at the request of Mr. McMuldroch and had no knowledge of Morville or the off-shore structure, nor did she ever attend any directors’ meetings or vote and thereby exercise what power she had. The only other director of APCI was Morville itself, the company owned for the benefit of the Bamber and McMuldroch families.
 Clearly, since Mr. Bamber was related to the Appellant by reason of controlling its parent company, BE Ltd., and hence is deemed not to have been at arm’s length with the Appellant under subsections 251(1) and (2) of the Act, and since Mr. Bamber clearly had de facto control of APCI and is thus considered not to have been at arm’s length with APCI in fact under paragraph 251(1)(c) of the Act; one may conclude that the Appellant and APCI were not at arm’s length under paragraph 251(1)(c) either.
 The Respondent also argued that, for the same reasons as those articulated above, the Appellant and APCI were related under paragraph 251(1)(a); however, as the Respondent conceded at the beginning of the trial that she was not relying on that specific provision, it is not necessary to comment further on that point.
 In conclusion, for the reasons stated above, I find that the Appellant was not at arm’s length with APCI, which brings into play both the transfer pricing rules of section 247 and the extended assessment period provisions of paragraph 152(4)(b) of the Act.
Limitation Period Issue
 There is no disagreement amongst the parties that the Minister reassessed the Appellant within the six-year total limitation period provided for in subparagraph 152(4)(b)(iii), above, and that the reassessments took place after the expiration of a period that was five years from the end of the taxation years in question, being the limitation period stated in the Treaty, as set out below. The Appellant takes the position that the Treaty limitation period applies to override the six-year limitation period prescribed by the Act, while the Respondent obviously takes the position that the Treaty limitation period does not apply and that, accordingly, the reassessments fall within the above-mentioned statutory period.
 The following table demonstrates the two limitation periods in question and the dates of the reassessments in issue:
Fiscal Year End
January 31, 1999
Sept. 20, 1999
January 31, 2004
Sept. 20, 2005
Sept. 16, 2005
January 31, 2000
July 31, 2000
January 31, 2005
July 31, 2006
May 18, 2006
January 31, 2001
August 9, 2001
January 31, 2006
August 9, 2007
May 18, 2006
 The question to be answered, then, is: Are the Minister’s reassessments statute-barred due to the application of the Treaty limitation period?
 The relevant articles of the Treaty are reproduced below:
Article I - Personal Scope
This Agreement shall apply to persons who are residents of one or both of the Contracting States.
Article II - Taxes Covered
1. This Agreement shall apply to taxes on income and on capital imposed on behalf of each Contracting State, irrespective of the manner in which they are levied.
2. There shall be regarded as taxes on income and on capital all taxes imposed on total income, on total capital, or on elements of income or of capital, including taxes on gains from the alienation of movable or immovable property, taxes on the total amounts of wages or salaries paid by enterprises, as well as taxes on capital appreciation.
3. The existing taxes to which the Agreement shall apply are, in particular:
(a) in the case of Canada: the income taxes imposed by the Government of Canada, (hereinafter referred to as "Canadian tax");
(b) in the case of Barbados:
(i) in the income tax;
(ii) the corporation tax;
(iii) the petroleum winning operations tax; and
(iv) the employment levy;
(hereinafter referred to as "Barbados tax").
4. This Agreement shall also apply to any identical or substantially similar taxes on income and to taxes on capital which are imposed by either Contracting State after the date of signature of this Agreement in addition to, or in place of, the existing taxes. The Contracting States shall notify each other of changes which have been made in their respective taxation laws.
Article III - General Definitions
1. In this Agreement, unless the context otherwise requires:
. . .
(e) the terms "enterprise of a Contracting State" and "enterprise of the other Contracting State" mean respectively an enterprise carried on by a resident of a Contracting State and an enterprise carried on by a resident of the other Contracting State;
. . .
(g) the term "tax" means Canadian tax or Barbados tax, as the context requires;
. . .
2. As regards the application of this Agreement by a Contracting State any term not otherwise defined shall, unless the context otherwise requires, have the meaning which it has under the laws of that Contracting State relating to the taxes which are the subject of this Agreement.
. . .
Article VII - Business Profits
1. The profits of an enterprise of a Contracting State shall be taxable only in that State unless the enterprise carries on or has carried on business in the other Contracting State through a permanent establishment situated therein. If the enterprise carries on or has carried on business as aforesaid, the profits of the enterprise may be taxed in the other State but only so much of them as is attributable to that permanent establishment.
. . .
Article IX - Associated Enterprises
(a) an enterprise of a Contracting State participates directly or indirectly in the management, control or capital of an enterprise of the other Contracting State, or
(b) the same persons participate directly or indirectly in the management, control or capital of an enterprise of a Contracting State and an enterprise of the other Contracting State,
and in either case conditions are made or imposed between the two enterprises in their commercial or financial relations which differ from those which would be made between independent enterprises, then any profits which would, but for those conditions have accrued to one of the enterprises, but, by reason of those conditions, have not so accrued, may be included in the profits of that enterprise and taxed accordingly.
2. Where profits on which an enterprise of a Contracting State has been charged to tax in that State are also included in the profits of an enterprise of the other Contracting State and taxed accordingly, and the profits so included are profits which would have accrued to that enterprise of the other State if the conditions made between the enterprises had been those which would have been made between independent enterprises, then the first-mentioned State shall make an appropriate adjustment to the amount of tax charged on those profits in the first-mentioned State. In determining such an adjustment due regard shall be had to the other provisions of this Agreement in relation to the nature of the income.
3. A Contracting State shall not change the profits of an enterprise in the circumstances referred to in paragraph 1 after the expiry of the time limits provided in its national laws and, in any case, after five years from the end of the year in which the profits which would be subject to such change would have accrued to an enterprise of that State. This paragraph shall not apply in the case of fraud, wilful default or neglect.
. . .
Article XXVII - Mutual Agreement Procedure
1. Where a resident of a Contracting State considers that the actions of one or both of the Contracting States result or will result for him in taxation not in accordance with this Agreement, he may, without prejudice to the remedies provided by the national laws of those States, address to the competent authority of the Contracting State of which he is a resident an application in writing stating the grounds for claiming the revision of such taxation. To be admissible, the said application must be submitted within two years from the first notification of the action which gives rise to taxation not in accordance with this Agreement.
2. The competent authority referred to in paragraph 1 shall endeavour, if the objection appears to it to be justified and if it is not itself able to arrive at an appropriate solution, to resolve the case by mutual agreement with the competent authority of the other Contracting State, with a view to the avoidance of taxation not in accordance with this Agreement.
3. A Contracting State shall not, after the expiry of the time limits provided in its national laws and, in any case, after five years from the end of the taxable period in which the income concerned has accrued, increase the tax base of a resident of either Contracting State by including therein items of income which have also been charged to tax in the other Contracting State. This paragraph shall not apply in the case of fraud, wilful default or neglect.
4. The competent authorities of the Contracting States shall endeavour to resolve by mutual agreement any difficulties or doubts arising as to the interpretation or application of this Agreement. In particular, the competent authorities of the Contracting States may consult together to endeavour to agree:
(a) to the same attribution of profits to a resident of a Contracting State and its permanent establishment situated in the other Contracting State;
(b) to the same allocation of income between a resident of a Contracting State and any associated person provided for in Article IX.
5. The competent authorities of the Contracting States may consult together for the elimination of double taxation in cases not provided for in this Agreement.
. . .
Article XXX - Miscellaneous Rules
1. The provisions of this Agreement shall not be construed so as to restrict in any manner any exclusion, exemption, deduction, credit, or other allowance now or hereafter accorded
(a) by the laws of one of the Contracting States in the determination of the tax imposed by that Contracting State, or
(b) by any other agreement between the Contracting States.
2. Nothing in this Agreement shall be construed so as to prevent Canada from imposing its tax on amounts included in the income of a resident of Canada according to section 91 of the Canadian Income Tax Act.
3. This Agreement shall not apply to companies entitled to any special tax benefit under the Barbados International Business Companies (Exemption from Income Tax) Act, Cap. 77 or to companies entitled to any special tax benefit under any similar law enacted by Barbados in addition to or in place of that law.
4. The competent authorities of the Contracting States may communicate with each other directly for the purpose of applying this Agreement.
 The Respondent argues that Articles IX(3) and XXVII(3) of the Treaty provide for a five-year limitation period and further states that the question of the applicability of the same provisions was brought by counsel for the Appellant before Rip C.J. of the Tax Court of Canada in the matter of Sundog Distributing Inc v. R., and that Rip C.J. found that neither limitation period in those articles applied. Counsel for the Appellant advised the Court that the decision of Rip C.J. has been appealed to the Federal Court of Appeal, which has yet to hear the appeal, and, accordingly, counsel suggested that I should consider waiting for that court’s decision.
 In summary, Rip C.J. decided that the five-year limitation period in Article IX(3) did not apply since the wording of paragraph IX(1) requires that, for its application, there must be “an enterprise of a Contracting State” from each of the states concerned, being Canada and Barbados, and that, since the appellant therein was a Barbados International Business Company (“IBC”), which Article XXX(3) specifically excludes from the application of the Treaty, it could not as an IBC, be “an enterprise of a Contracting State” for the purposes of Article IX of the Treaty.
 Frankly, I am in full agreement with the decision of Rip C.J. on this issue and adopt his reasoning on the non-applicability of the limitation period in Article IX(3). The Appellant itself made no further argument on this provision.
 I should also like to add however, that it is important to note that the parties in that case, like those in this matter, did not dispute that Article IX of the Treaty is a provision that deals specifically with transfer pricing. By its very terms, Article IX requires that an appropriate adjustment to taxes be made by the state that did not make the transfer pricing adjustment, and thus it is also clear that the limitation period in Article IX(3) is intended to ensure that the state charged with making an adjustment to taxes charged by it under Article IX(2) does not have an open-ended obligation to do so ‑ instead it is only subject to this requirement for a period of five years. In summary, we have an article that specifically deals with transfer pricing issues, setting out an obligation to adjust taxes, when transfer pricing adjustments occur in order to avoid double taxation, and containing its own limitation period. In this context alone, one must question the applicability of another, more general limitation period found in a different part of the Treaty, such as Article XXVII(3), especially where that other limitation period is a similar one. This interpretation is bolstered further when one considers Article IX of the Treaty together with the related OECD Commentary. As explained in paragraph 5 of the OECD Commentary on Article IX, when one state adjusts a taxpayer’s income under the transfer pricing provisions in paragraph IX(1), paragraph IX(2) mandates that the second state “make an appropriate adjustment so as to relieve the double taxation.” Moreover, as stated at paragraph 11 of the Commentary:
11. If there is a dispute between the parties concerned over the amount and character of the appropriate adjustment, the mutual agreement procedure provided for under Article 25* should be implemented; the Commentary on that Article contains a number of considerations applicable to adjustments of the profits of associated enterprises carried out on the basis of the present Article (following, in particular, adjustment of transfer prices) and to the corresponding adjustments which must then be made in pursuance of paragraph 2 thereof (see in particular paragraphs 10, 11, 12, 33, 34, 40 and 41 of the Commentary on Article 25).
*While Article 25 of the model convention corresponds with Article XXVII of the Barbados Treaty, the comments about Article 25 and its ties to Article 9 should not be interpreted as applying to paragraph XXVII(3) of the Treaty as that provision is a unique addition and is not a part of the model convention.
 Finally, even if one looks just at the text of the Treaty, one finds further support for this interpretation, as, with respect to transfer pricing, Article IX of the Treaty does not reference the general limitation period found in paragraph XXVII(3) of the Treaty; instead, the only limitation period directly applicable to transfer pricing is the one found in the transfer pricing article itself.
 The decision of the Supreme Court of Canada in Crown Forest Industries Ltd. v. Canada, cited by the Appellant, emphasizes the necessity of looking at the language used in the Treaty provisions together with the parties’ intention in drafting those provisions, which includes having reference to extrinsic material such as general models of international taxation conventions. When the Treaty is read as a whole and the OECD Commentary illuminating Article 9 of the model convention and tying it to Article 25 thereof is read with it, one can see that it would be redundant to apply paragraph XXVII(3) of the Treaty to transfer pricing cases. Instead, the more logical conclusion is that the paragraph was added by Canada and Barbados as a general provision intended to catch instances not addressed in other parts of the Treaty. For example, it may be the case that profits which include the same items of income stand to be taxed in both states, i.e., where a taxpayer has a permanent establishment in the other contracting state, and Article VII therefore applies. This conclusion is reinforced by the fact that Article VII does not have its own specific limitation provision.
 Notwithstanding the above argument as to why I believe it is clear that the limitation provision in paragraph XXVII(3) does not apply, in Sundog its applicability was analyzed in a different context and also found not to apply.
 In paragraph 46 of Sundog, Rip C.J. stated:
46 I do not agree with the appellant’s submissions. First of all, Article XXVII(3) prohibits a State from increasing a person’s tax base by including items of income already charged to tax in the other State; it is not income generally. Secondly, Barbados has never charged tax to the appellant on its income. Thirdly, the word “income” and “revenu” (in the French version of Article XXVII(3)) are modified by the words “items of” in English and “éléments de” respectively. It is not income alone, or the quantum of the income, that is addressed in Article XXVII(3); it is a description of what part of income has entered into an account or a “partie constitutrice d’une chose”, a constructive or essential part of a thing that is being charged to tax. It is not the general description of income but what the income is composed of, what the income is, that interests Article XXVII(3). The income may be a dividend, interest, sales, professional fees or management fees, among others.
 Rip C.J. went on to find that different types of income were involved in that Canada taxed the appellant on management fees covered by Article XIII of the Treaty from sources in Barbados or on business profits covered by Article VII, while Barbados did not charge the appellant any taxes on these items as it had no permanent establishment in Barbados and was thus exempt from tax there. Canada was actually charging the appellant tax on management fees that it stated ought to have been paid by the Barbados companies, i.e., Sun Island Optics and Sun Island, but were not, and was, in fact, making an income adjustment for fees the two Barbados companies ought to have paid.
 In paragraph 48, Rip C.J. summarized the matter as follows:
48 . . . Again, the appellant was not charged tax on management fees or on any type of income by Barbados; the items of income charged to tax against the appellant by Canada are different items of income than were charged to tax to Sun Island Optics and Sun Island by the Government of Barbados.
 The Appellant argues that its counsel’s argument in Sundog ‑ counsel for the Appellant herein also having represented the appellant in that case ‑ was the correct view, in that it is not necessary to determine the nature of the income given that Article XXVII applies to income generally, and that both governments would be taxing the Appellant since Canada is trying to assess tax on the same income – i.e. just as Barbados had assessed Sun Island Optics and Sun Island in Sundog. In other words, the Appellant claims that Canada, in assessing the Appellant for tax on the adjustment made, has increased the tax base of the Appellant by including income charged to tax in Barbados.
 Frankly, the Appellant’s argument in Sundog ignores the obvious wording of paragraph XXVII(3), which refers to “items of income” and not income generally. Rip C.J.’s interpretation of “items of income” as consisting of specific items of revenue is also supported by other provisions of the Treaty itself, including paragraph VII(6) which begins “Where profits include items of income . . .”. In my view, Rip C.J. was correct in his interpretation of “items of income”.
 The Appellant also argues in the case at bar that we are dealing with the same items of income. In effect, the Appellant argues that the set-up fees included in income in Barbados constitute the same income as the business expense deducted by the Appellant in Canada. Frankly, I have serious difficulty with regard to this argument for two reasons. First, the Appellant took a business expense deduction for the set-up fees paid to APCI. This is an item of expense, not income, notwithstanding that both are of the same quantum. The most that can be said is that the expense and the item of income both affect the net business income of both parties; but this is clearly not what is contemplated by the wording of paragraph XXVII(3), which, in effect, contemplates the same items of income being taxed in both states. It is, in fact, abundantly clear that paragraph XXVII(3) can only deal with a situation where both contracting states exercise jurisdiction to tax the same items of income, presumably items not otherwise covered in the Treaty, resulting in the obligation of both contracting states under paragraph XXVII(2) to endeavour to resolve the issue with a view to avoiding double taxation. Secondly, unlike the predecessor transfer pricing rules found in former subsection 69(2) of the Act ‑ applicable to taxation years that began prior to 1998 ‑ subsection 247(2) works so as to adjust the income of the taxpayer up or down, not to reduce the expense amount taken as a deduction from the payments made to a non-arm’s length party. This is, however, only an adjustment to income. It is not an item of income per se and by no means is this upward adjustment of income in the case at bar also taxed as an additional item of income in Barbados. In short, the level of deductions in Canada or the level of income in Barbados is not technically altered, notwithstanding that the substantive effect would be that part of the Appellant’s deduction of set-up fees paid to APCI would be effectively neutralized by the adjustment.
 I should also add that while the bulk of the Appellant’s argument dealt with his disagreement with Rip C.J.’s interpretation of the phrase “items of income” in paragraph XXVII(3), the Appellant seems to have ignored the other crucial requirement of that Article, namely that the items of income must have also “been charged to tax in the other Contracting State”. Accordingly, even if the decision of Rip C.J. on this issue is overturned on appeal or the items of income in this case are found to be the same, the Appellant must still demonstrate that such items of income were charged to tax in Barbados.
 The Appellant takes the position that since APCI paid the 2.5% levy or tax pursuant to the IBC Exemption From Income Tax Act, it accordingly paid tax on that item of income in Barbados. There is no dispute that APCI, as an IBC, was exempt from tax pursuant to the Barbados Income Tax Act. The question then becomes whether this “tax” paid by APCI was the tax contemplated by the Treaty in paragraph XXVII(3).
 Article III, subparagraph (1)(g) of the Treaty states that “the term ‘tax’ means Canadian tax or Barbados tax, as the context requires.” Article II further defines the term “Barbados tax”:
(3) The existing taxes to which the Agreement shall apply are, in particular:
. . .
(b) in the case of Barbados:
(i) the income tax;
(ii) the corporation tax;
(iii) the petroleum winning operations tax; and
(iv) the employment levy;
(hereinafter referred to as "Barbados tax").
 Although the terms in clauses 3(b)(i) to (iv) above are not further defined in the Treaty, paragraph 2 of Article III elaborates further:
(2) As regards the application of this Agreement by a Contracting State any term not otherwise defined shall, unless the context otherwise requires, have the meaning which it has under the laws of that Contracting State relating to the taxes which are the subject of this Agreement.
 To begin, while IBCs benefit from significantly reduced tax rates, they are still taxed. Subsection 10(1) of the IBC Exemption From Income Tax Act states:
10(1) Subject to this section and section 11, in lieu of tax at the rate specified under the Income Tax Act, there shall be levied and paid to the Commissioner of Inland Revenue, in respect of the income year 1991 and in each subsequent income year of an international business company, a tax on the profits and gains of the company at the following rates . . .
 This leaves open the question of whether or not the tax in question falls under (or is identical or substantially similar to) one of the categories listed in Article II, subparagraph (3)(b) of the Treaty. The Barbados Income Tax Act clarifies this matter in section 3, where it states:
Charge to Tax
3. (1) Except as provided by subsection (2), every person who has a taxable income for an income year shall pay in accordance with this Act an income tax calculated in accordance with this Act on his taxable income for that income year.
(2) A company that has a taxable income in any income year shall, instead of paying income tax under subsection (l), pay a corporation tax on that income in respect of that year calculated in accordance with this Act.
 Beginning with subsection 3(1), one can see that income tax is payable unless subsection 3(2) applies. Under subsection 3(2), a company is required to pay a “corporation tax” instead of paying “income tax” under subsection (l). At this point, one can conclude that companies pay “corporate tax” and individuals pay “income tax”. The term “company” is defined in both the Barbados Income Tax Act and the IBC Exemption From Income Tax Act as including a body corporate however and wherever incorporated – a definition that includes IBCs.
 This interpretation is confirmed by the definition of “tax” found in section 85 of the Barbados Income Tax Act, which states:
(i) in relation to a person liable to tax under subsection (1) of section 3, the income tax imposed by that subsection,
(ii) in relation to a company the corporation tax imposed by section 3(2).
 As seen above, under section 10 of the IBC Exemption From Income Tax Act, a special tax is levied on IBCs in lieu of tax at the rate specified under the Barbados Income Tax Act.
 At this point, it must be noted that the first iteration of the IBC Exemption From Income Tax Act was enacted in 1965, long before the Treaty entered into force on December 22, 1980. One can thus conclude that the drafters of the Treaty were aware of the nature of IBCs when drafting the articles of the Treaty. This conclusion is reinforced by the existence of paragraph XXX(3), which expressly excludes companies entitled to any special tax benefit under the IBC Exemption From Income Tax Act – i.e., IBCs – from the application of the Treaty.
 In light of this, it is possible to conclude that the tax provided for in the IBC Exemption From Income Tax Act is neither the “income tax” ‑ which arguably only applies (as a phrase) to individuals, nor the “corporation tax” prescribed in subsection 3(2) and whose rate is specified in section 43, as the tax in section 10 of the IBC Exemption From Income Tax Act is levied in lieu of tax at the rate specified under the Barbados Income Tax Act.
 In summary, the tax levied under the IBC Exemption from Income Tax Act does not count as a “tax” under the Treaty. Barbados and Canada chose to establish an inclusive, fixed list of taxes that would be considered “Barbados tax”, and that list does not include the tax levied under the IBC Exemption From Income Tax Act. In other words, although the Appellant was subject to a tax, it was not subject to “tax” either as that term is defined in the Treaty, or within the context of the Treaty requirements.
 Based on the above analysis, I conclude that the Minister’s assessments are not statute-barred by paragraph XXVII(3) of the Treaty as the Appellant’s income was not subject to “tax”, nor was the inclusion in income an “item of income” (or at least not one that matches the items of income included in APCI’s income), and accordingly Article XXVII does not apply. Consequently, the Minister issued the Notices of Reassessment within the aggregate six-year period contemplated by paragraph 152(4)(b) of the Act and is not statute-barred from proceeding with those reassessments.
Transfer Pricing Issue
 The transfer pricing adjustments made by the Minister in the reassessments referred to above are, of course, the main issue to be determined in this appeal. Given the finding that the Appellant and APCI were not dealing at arm’s length, the precondition to the applicability of subsection 247(2) of the Act has been met. Before proceeding with a detailed analysis of the transfer pricing rules, I propose to deal with the most serious factual disputes between the parties relating to the comparability factors relevant to determining the transfer pricing issue, namely, the matters of what services were being provided by APCI to the Appellant and at what price. These are crucial questions to be determined since the parties have made different assumptions on these issues in the transfer pricing methodologies used by them and in presenting argument on why the methodologies used by the other party were incorrect.
 There is, of course, no dispute that APCI was a provider of services to the Appellant and that the totality of the services in question consisted of both the set‑up function previously performed by the Appellant in Canada and the provision of software development and maintenance and website development and maintenance services, as described in the Functional Analysis and the annual contracts referred to in the summary of the facts above. Part 2 of those contracts related to set-up and Parts 1 and 3 had to do with software development and maintenance and website development and maintenance. The factual dispute between the parties is not so much a dispute as to the categorization of services described in the above-mentioned three Parts as a dispute as to the allocation of the prices amongst them and whether APCI in fact performed all of the set-up functions.
 The Minister’s position on these issues is as follows:
1. At least a portion of the square inch charge the Appellant argues was compensation for the non-set-up services provided by APCI, namely the Part 1 and 3 services, was itself for set-up fees. In other words, the 22.01% of the total fees paid to APCI by the Appellant representing the annual square inch fee was wholly or partly for the set-up services as well.
2. The Appellant itself continued to perform some of the set-up services and hence the basic set-up fee paid to APCI was overstated for this reason as well.
 The only assumptions pertaining to these factual disputes are the following found in paragraph 17 of the Minister’s Reply:
17. In determining the Appellant’s tax liability for the 1999, 2000, and 2001 taxation years, and with respect to the matters at issue in this appeal, the Minister made the following assumptions of fact:
. . .
p) . . . The Appellant thereby moved the “set-up” function department, which it had previously solely operated, to APCI;
q) Starting in January of 1997, the Appellant outsourced its “set-up” functions to APCI;
. . .
s) The Appellant determined the prices charged by APCI for the services rendered to the Appellant.
. . .
u) The Appellant was able to perform the “set-up” functions in Canada and, on occasions when APCI’s office was shut down or encountered complications, the Appellant did so.
 The Respondent’s position on the first issue, namely that the set-up fee included some or all of the square inch fee, is chiefly premised on the wording of the Functional Analysis and annual contracts.
 The Respondent, in cross-examination of each of Mr. and Mrs. Bamber, Mr. McMuldroch, and the Appellant’s expert witness, Mr. Wall, referred to the final paragraph of the Functional Analysis (prepared jointly by Mr. McMuldroch and Mrs. Bamber), which the Respondent contends is evidence that the set-up fee included the square inch charge and hence would have increased the set-up fee from $67 to $93 in the 1998 fiscal year. That paragraph is reproduced below:
In summary, APCI, Inc. charged approximately $67.00 Canadian for their setup fee to Alberta Printed Circuits Ltd. In addition, APCI, Inc. charged an annual performance fee to Alberta Printed Circuits known as the “square inch charge”. In 1998 that additional “square inch charge” was $300,000 Canadian. There were approximately 12,000 setups done by APCI, Inc. in 1998. That represents, on average, an additional fee of $26 Canadian per setup making the total setup fee $93.00 Canadian. One of APCI, Inc.’s comparable competitors, Star Electronics, Corp. charged a setup fee of between $100.00 and $125.00 US at that time. Clearly, the fee charged by APCI, Inc. is reasonable and meets the “arm’s length” pricing objective under transfer pricing law.
 Both Mr. and Mrs. Bamber denied that any portion of the square inch fee was for set-up and, in fact, on re-examination counsel for the Appellant brought the Court’s attention to the first paragraph on page 10 of that analysis, which states:
The only other amount that was paid to APCI, Inc. by Alberta Printed Circuits Ltd. other than the setup and setup-related charges, was the annual square inch fee. This fee was based on a percentage of the total manufactured square inches and served the purpose of a bonus in part for APCI, Inc. providing software support, customer relations (the Barbados staff handled necessary customer calls related to problems with data formats and information that they had received from Alberta Printed Circuits Ltd. and the telephone charges were very expensive), web site maintenance, etc. (these other services were listed in the annual agreement which CRA has a copy of). (see Appendix A for additional information).
 While one can appreciate the Respondent’s position (based on the last paragraph of the Functional Analysis), it is absolutely clear that the first sentence, above, states: “The only other amount that was paid to APCI, Inc. by Alberta Printed Circuits Ltd. other than the setup and setup related charges, was the annual square inch fee.” These words are clearly inconsistent with the Respondent’s position. Moreover, as Mr. Wall suggested during his cross‑examination, in his view the last paragraph of the Functional Analysis addressed the overall assumption made by the Canada Revenue Agency (“CRA”) that “the amount paid by APCI was unreasonable” in total, and clearly the authors of the Functional Analysis were attempting to point out that even the total amount would have been considered reasonable if one adopted the Respondent’s position.
 The Respondent also relied on the wording of the annual contracts, particularly the contract applying to the fiscal period February 1, 1999 to January 31, 2000, which described the prices under the Proto 1 (P1) service rate structure as follows:
Initial set up fee $67 per order + $0.2285 per square inch
 In fact, the juxtaposition of the set-up fee and the square inch fee appears in their first contract, which was for the period February 1, 1997 to January 31, 1998, and in the third annual contract, for the period February 1, 1999 to January 31, 2000, as well, while the juxtaposition of those two rates does not appear in the final contract, commencing February 1, 2000, where reference to a reduced square inch rate is found in a separate paragraph. Having regard to this inconsistency, one must look to other evidence to ascertain the intention of the parties, and, in this regard, I am in agreement with the Appellant that, as the base rate was applied on a transaction by transaction basis and the square inch fee was only due at the end of the year, preference must be given to consideration of the transactional fees. Moreover, as the Respondent herself pointed out, the square inch fee appeared to have been reduced over the years, particularly when the non-set-up services performed by APCI were reduced on Mr. McMuldroch’s return to Calgary in 1999, which, frankly, supports the position that the square inch or bonus fee only applied to the Part 1 and Part 3 services described in the annual contracts referred to above. The Respondent’s argument that the decrease of $50,000 in the total square inch fees in 2000, when Mr. McMuldroch returned to Calgary, represents the entire value of the services for Parts 1 and 3 and hence the balance must be considered as applying to the set-up fee, is not, per se, convincing as it ignores the terms of the annual contract, which set the price for such services far above $50,000. Moreover, this does not address the issue of why those services should not have been valued at that amount in the first place, nor does it deal with the fact that a reduction of $50,000 for that year, when Mr. McMuldroch spent only 4½ out of 12 months in Barbados, would suggest an annualized rate for the year of about $80,000.
 With respect to the second issue whether the Appellant performed some of the set-up functions, the Respondent also submitted that the first three set-up steps described by Mr. Wall at page 9 of his report on the set up were clearly suggestive of the fact that the Appellant retained responsibility for part of the set‑up function and that Mr. Wall erred in assuming APCI was responsible for the entire set‑up function. The excerpts referred to by the Respondent are as follows:
Below are excerpts from the Functional Analysis that briefly describe the steps of the Set up activities, following the natural sequence of events for a customer order.
1. AP Circuits received data (described above) from customers via email attachment, file transfer protocol (FTP) used to transfer files over the internet, or the Bulletin Board Service (BBS).
2. AP Circuits collected all of the email attachments and placed the files in an area where they could be processed by a computer program that assigned job numbers to each data set.
3. AP Circuits checked that all of the required data was present in each file before sending it to APCI.
4. APCI was responsible for performing the formatting and panelizing (described below).
5. APCI transmitted the finalized outputs to AP Circuits via the internet or telephone line.
6. AP Circuits checked the outputs for accuracy and then sent the data via an internal network to the drill and photoplotter, which was the beginning of the manufacturing process for a circuit board.
7. APCI provided email and support desk services for AP Circuits’ customers.
 The description of the set-up function through the testimony of Mr. McMuldroch, which I found to be credible on this subject in particular, clearly suggests that the set-up stage commenced after the third step, above. The evidence was clear that the data obtained by the Appellant through those three steps would be transmitted to APCI electronically. APCI would then proceed with its steps of validation, formatting, and panelizing as above described. It should also be noted that another of the Appellant’s witnesses, a Mr. James Malcolm Godfrey, testified that he was a manufacturer of circuit boards from 1979 to 2008 under the name Western Index Manufacturing and performed all roles, including set up and manufacturing, and he described the set‑up steps in similar manner as Mr. McMuldroch and the Appellant did, i.e., verification, formatting, and panelizing.
 Accordingly, I accept Mr. Wall’s evidence that he was blending both the pre- and post-set-up steps and the set-up phase itself in his explanation describing the relevant parties’ responsibilities and that he too considered the set-up function as starting at step four of the steps listed by him. Frankly, the first three steps are clearly consistent with the “natural sequence of events for a customer order” referred to in Mr. Wall’s report.
 The Respondent also suggested that, in light of Mr. Wall’s step 7, above, and a reference in the Functional Analysis to APCI providing customer support as part of the services it rendered for the square inch fees at least part of the square inch fees were for set-up. Again, the evidence clearly establishes that, as a part of the set-up obligation, APCI would contact the customers’ engineers in the event that anomalies were detected any anomalies in the data supplied, and this was what, in my view, was being described in Mr. Wall’s step number 7, above. It is also reasonable to assume that the customer support referred to in the relevant paragraph of the Functional Analysis clearly refers to APCI’s services described in Parts 1 and 3 of the annual contracts, that is to say, a different kind of customer support.
 With respect to the second issue, i.e., whether part of the set-up function was performed in Canada by the Appellant, the Respondent suggests that, since an employee of the Appellant checked for errors in the set-up information electronically transmitted by APCI to the Appellant – a function carried out in particular by Mr. Wu prior to his leaving the employ of the Appellant in December 1997, but subsequently by other employees of the Appellant ‑ that this was part of the set-up function being performed in Canada. With respect, the Respondent’s own witness, Ms. Lewinske, testified that this was not a step that even existed before she moved to Barbados, so it could not have been part of the previous set-up function. The explanation given for this step by the Appellant was that it was mainly to check to ensure that the set-up information transmitted from Barbados to Canada was not corrupted during transmission (a quality control check) and I find this to be very credible.
 As for the Respondent’s assumption that the Appellant was able to and did perform set-up services from time to time, the only evidence that someone may have been skilled enough to perform some of the set-up function for the Appellant after the move of that function to Barbados was again with respect to the services of Mr. Wu, an employee who was part of the set-up department before the Barbados move and the only member of the department who did not move to Barbados. The evidence of the Appellant was not only that he was the least experienced of the three set-up people and not able to perform all the set-up functions at the time of the move, but also that he ceased to be an employee of the Appellant in December 1997, that is, before the years in dispute, so he could not have performed that function on behalf of the Appellant in any event. There was absolutely no evidence of anyone else having that ability after Mr Wu’s departure and the only other evidence of anyone at the Appellant’s premises being trained to perform set-up thereafter was that, on Mr. McMuldroch’s return in June of 1999, the Appellant started to look to hire people to be trained in set‑up. Again, it took anywhere from six months to a year for such persons to become totally dependable, and in the meantime set-up remained in Barbados. In any event, the fact that someone could have done set-up does not mean that they did, and the evidence was that APCI continued to do the set-ups until the arrangement was terminated at the end of 2000.
 In my view, the evidence establishes that the base rate fee was the only fee paid for set-up services. That evidence was not satisfactorily rebutted by the Respondent, and all of the set-up function was performed in Barbados by APCI during the taxation years in issue (other than the P2 set-up, which was done by Mr. McMuldroch after his return, as he was the only one capable of doing such set-ups, and for which APCI received no compensation after Mr. McMuldroch’s return to Canada). Having said this, of course, the issue still remains whether the fees paid for the set-up service or the other services were at reasonable, arm’s length prices. I must also conclude that Mr. Wall did not err in using these assumptions for the purposes of his comparable uncontrolled price (“CUP”) analysis and that he did not recharacterize the transaction as the Respondent stated, which will be discussed further below.
 While there were other, comparable disputes as to the relative risks assumed and the intangibles owned by the parties, these are more relevant to the application of the transactional net margin method (“TNMM”), to be discussed later; to the extent that they are the factors to consider, my comments thereon apply equally in the context of other pricing methods considered, particularly the CUP analysis.
Transfer Pricing Law and Rules
 Paragraphs 247(2)(a) and (c) are the relevant provisions of the Act dealing with the transfer pricing issue in this matter and are again reproduced below:
247(2) Where a taxpayer or a partnership and a non-resident person with whom the taxpayer or the partnership, or a member of the partnership, does not deal at arm’s length (or a partnership of which the non-resident person is a member) are participants in a transaction or a series of transactions and
(a) the terms or conditions made or imposed, in respect of the transaction or series, between any of the participants in the transaction or series differ from those that would have been made between persons dealing at arm’s length, or
. . .
(c) where only paragraph 247(2)(a) applies, the terms and conditions made or imposed, in respect of the transaction or series, between the participants in the transaction or series had been those that would have been made between persons dealing at arm’s length, or . . .
There is no dispute that paragraphs (b) and (d) of subsection 247(2) of the Act have no application.
OECD Convention and Guidelines
 The Federal Court of Appeal has said in SmithKline Beecham Animal Health Inc. v. R. that:
8 It appears to be common ground that the OECD Guidelines inform or should inform the interpretation and application of subsection 69(2) of the Income Tax Act. . . .
 Subsection 69(2) was replaced by subsection 247(2) of the Act for taxation years beginning in 1998, which covers all the taxation years under appeal here. However, as the Federal Court of Appeal confirmed in paragraph 12 of General Electric Capital Canada Inc. v. Canada, “. . . there is no meaningful difference between paragraphs 247(2)(a) and (c) and subsection 69(2) of the Act”.
 The CRA, of course, relies on the OECD Transfer Pricing Guidelines for Multinational Enterprises and Tax Administrations in assessing: Information Circular 87-2R, International Transfer Planning, dated September 27, 1999, which replaced former IC087-2 dated February 2, 1987.
 Paragraphs 247(2)(a) and (c), like former subsection 69(2), is analogous to Article 9(1) of the OECD Model Double Taxation Convention on Income and Capital. The OECD issued a Commentary on transfer pricing in 1979, which was updated in 1995. There was a further update in 2010, but, since this update is well beyond the taxation years in issue, I will refer only to the applicable 1995 Commentary.
 It is clear that both the CRA Information Circular 87-2R, paragraphs 7, 8, and 28, and the OECD Commentary on Article 9 rely on the arm’s length principle to determine prices, referred to as “transfer prices”, which are charges for goods and services traded across international borders between related parties.
 According to the OECD Commentary, such parties may adopt transfer prices for reasons other than to minimize tax, but, regardless of the reason, intra-group transfers which are not carried out at arm’s length prices will likely result in profits being shifted from one country to another.
 The underlying policy concern behind the transfer pricing rules is, of course, leakage from the Federal Treasury due to profits being shifted from one country to another or, expressed in more conventional terms, the object is to ensure that parties not at arm’s length report substantially the same amount of income in the jurisdiction in which they are located as would parties dealing at arm’s length.
 A hierarchy of methods that can be used to determine the transfer price is set out in the OECD Guidelines. There are three “traditional transaction methods”:
· The comparable uncontrolled price method (CUP);
· The resale price method;
· The cost plus method; and
There are also two “transactional profit methods”:
· The profit split method; and
· The transactional net margin method (TNMM).
 Each of the methods generally involves a comparison of prices or margins between non-arm’s length parties on cross-border transactions (“controlled transactions”) with prices or margins on similar transactions between arm’s length parties (“uncontrolled transactions”).
 The above OECD methods either as described by the OECD itself – whose descriptions were also reproduced (in part) by Rip A.C.J., as he then was, at paragraph 63 of GlaxoSmithKline v. The Queen, (GlaxoSmithKline) – or as paraphrased by this Court are as follows:
Traditional Transaction Methods
(a) The comparable uncontrolled price method offers the most direct way of determining an arm's length price. The transfer price is set by reference to comparable transactions between a buyer and a seller who are not associated enterprises. Uncontrolled sales may include sales by a member of an MNE to an unrelated party and sales to a member of an MNE by an unrelated party as well as sales in which the parties are not related to each other or to the MNE (though they may themselves be members of other MNEs). Uncontrolled sales are, in short, sales in which at least one party to the transaction is not a member of the taxpayer's affiliated group, but they would include only bona fide transactions and not sales unrepresentative of the market, for example made in a limited quantity at unrealistic prices to an unrelated buyer, for the purpose of establishing an arm's length price on a larger transaction. The method requires the uncontrolled transactions to be carefully reviewed for comparability with controlled transactions. . . .
(b) The resale price method begins with the price at which a product which has been purchased from a related seller is resold to an independent purchaser. This price is then reduced by an appropriate mark-up representing the amount out of which the reseller would seek to cover his costs and make a profit. What is left after subtracting the mark-up can be regarded as an arm's length price of the original sale. This method is probably most useful where it is applied to marketing operations.
(c) This method [cost-plus] of estimating an arm's length price is based on the supplier's cost to which an appropriate profit-mark-up is added. It is a method that raised problems both as regards assessing costs . . . and the appropriate mark-up for profit . . . and is likely to be appropriate as a deterring criterion mostly in specific situations . . . though it may also be useful as a means of verifying provisionally acceptable prices after other methods have been applied. . . . . . .the cost plus method may be helpful in estimating an arm's length price, when semi-finished products are sold between related parties. . . .
Transactional Profit Methods
(d) The profit-split method of estimating an arm's length price identifies the combined profit to be split for the associated enterprises from a controlled transaction (or controlled transactions that it is appropriate to aggregate) and then splits those profits between the associated enterprises based upon an economically valid basis that approximates the division of profits that would have been anticipated and reflected in an agreement made at arm's length. Where possible, this economically valid basis may be supported by independent market data (e.g. division of profits observed in uncontrolled joint-venture agreements). Most often, however, it will be supported by internal data. The splitting factor should reflect the respective contributions of the parties to the creation of income from the controlled transaction and be reasonably independent from transfer pricing formulation. This means that it should, to the greatest extent possible, be based on objective data (such as sales to unrelated parties), rather than on data relating to the remuneration of controlled transactions (such as sales to associated enterprises).
(e) The transactional net margin method examines a net profit indicator, i.e. a ratio of net profit relative to an appropriate base (e.g. costs, sales, assets), that a taxpayer realises from a controlled transaction with the net profit earned in comparable uncontrolled transactions. Thus, a transactional net margin method operates in a manner similar to the cost plus and resale price methods, except it compares the net profit and not the gross profit on resale or gross mark up on cost. Because of this, the net profit indicator of the taxpayer from the controlled transaction should ideally be established by reference to the net profit indicator that the same taxpayer earns in comparable uncontrolled transactions (in other words, by reference to “internal comparables”). A functional analysis of the controlled and uncontrolled transactions is required to determine whether the transactions are comparable and what adjustments may be necessary to obtain reliable results as there are various factors other than products and functions that can significantly influence net profit indicators. Functional comparability is generally found to be of greater importance than product comparability in applying the transactional net margin method. This method is most useful where there is no or limited publicly available reliable gross margin information on third parties.
Depending on the facts and circumstances of the case and in particular on the effect of the functional differences on the cost structure and on the revenue of the potential comparables, greater attention must be paid to the comparability requirements. For example, net profit indicators can be less sensitive than gross margins to differences in the extent and complexity of functions and to differences in the level of risks (assuming the contractual allocation of risks is arm’s length), but more sensitive than the cost plus or resale price methods to differences in capacity utilization because differences in the levels of absorption of indirect fixed costs if not reflected in price differences. Moreover, net profit indicators may be directly affected by such forces operating in the industry as follows: threat of new entrants, competitive position, management efficiency and individual strategies, threat of substitute products, varying cost structures (as reflected, for example, in the age of plant and equipment), differences in the cost of capital (e.g. self-financing versus borrowing), and the degree of business experience (e.g. whether the business is in a start-up phase or is mature). Each of these factors in turn can be influenced by numerous other elements. For example, the level of the threat of new entrants will be determined by such elements as product differentiation, capital requirements, and government subsidies and regulations. A final important aspect of comparability is measurement consistency. The net profit indicators must be measured consistently between the associated enterprise and the independent enterprise. In addition, there may be differences in the treatment across enterprises of operating expenses and non‑operating expenses affecting the net profits such as depreciation and reserves or provisions that would need to be accounted for in order to achieve reliable comparability. Finally, a transactional net margin method is unlikely to be reliable if each party to a transaction makes valuable, unique contributions.
 The test to be used in determining whether the Appellant’s price was appropriate for the purposes of the transfer pricing provisions was stated in the recent Federal Court of Appeal decision of GlaxoSmithKline Inc. v. The Queen, (Glaxo) at paragraph 69, where Nadon J.A. relied upon the Exchequer Court’s decision in Gabco Limited v. Minister of National Revenue, in which Cattanach J. said:
It is not a question of the Minister or his [sic] Court substituting its judgment for what is a reasonable amount to pay, but rather a case of the Minister or the Court coming to the conclusion that no reasonable business man would have contracted to pay such an amount having only the business consideration of the appellant in mind.
 In applying that test, the Federal Court of Appeal stated in paragraph 73 of Glaxo:
73 In my view, the test set out in Gabco, supra, requires an inquiry into those circumstances which an arm’s length purchaser, standing in the shoes of the appellant, would consider relevant in deciding whether it should pay the price paid by the appellant . . .
That court further stated in paragraph 74:
74 . . . In other words, the test mandated by subsection 69(2) [now paragraphs 247(2)(a) and (c)] does not operate regardless of the real business world in which the parties to a transaction participate.
 What is clear from the analytical framework established by the Federal Court of Appeal in Glaxo is that the analysis is a strongly fact-driven one, carried out in the context of basic business principles, or the “real business world” as the Federal Court of Appeal described it.
 Accordingly, regardless of which of the different methodologies in the OECD hierarchy above described are used, the test requires a consideration of all relevant factors that a reasonable business person in the Appellant’s shoes would consider.
 It is important to note that factors or circumstances that exist solely because of the non-arm’s length relationship of the parties should not be ignored, otherwise the reasonable businessman would not be standing entirely in the Appellant’s shoes. This was demonstrated in Glaxo above.
 The OECD terms these relevant factors the “Comparability Factors” in its Guidelines and included the following categories of factors to consider:
1. The characteristics of property or services being purchased or sold
- differences may account for differences in value;
2. The functions defined by the parties to the transaction – i.e., who does what, taking into account assets used and risks assumed;
3. Contractual terms between the parties;
4. The economic circumstances of the parties - what markets do they operate in and what are the differences?
5. The business strategies pursued by the parties.
 The list, of course, is not intended to be exhaustive, as consideration of all relevant factors is mandated. In Glaxo, the Federal Court of Appeal ruled that the Tax Court had to consider a licensing agreement between the non-arm’s length parties as well as the contract between them. In General Electric, the Federal Court of Appeal confirmed that no error in law was made in taking into consideration that the Appellant in that case, as a subsidiary of its larger parent company, stood in the position of having an implicit guarantee by its parent of its bank debts.
 In short, all circumstances means “all” the circumstances an appellant finds himself in before a reasonable businessman steps into his shoes.
 In paragraphs 36 and 37 of IC 87-2R, the Canada Revenue Agency (the “Department”) states that “the arm’s length principle should ideally be applied on a transaction-by-transaction basis. Therefore, in establishing transfer prices, taxpayers should set prices separately for each transaction they enter into with a non‑arm’s length party. This separate determination usually provides the most reliable estimation of an arm’s length price.” However, the Department recognizes in paragraph 37 that in situations where transactions are so closely linked they cannot be evaluated adequately on a separate basis, it may be necessary to bundle transactions.
 The Department further recognizes that the bundling or unbundling of transactions for the purposes of determining an arm’s length price does not change the characterization of the underlying transactions.
 I take this to clearly mean it is preferable and more reliable to unbundle products or services when data is available to evaluate them on a separate basis; i.e., if one party sells apples and oranges and only buys apples from a non-arm’s length party, one can evaluate a separate transfer price for apples instead of leaving the two bundled. In our case, we can separate the set-up from the other services provided by APCI and from the manufacturing done by the Appellant, which allows us to compare the same service.
 It is understood that the selection of the most appropriate pricing method depends largely on the assessment of the comparability of transactions.
 The Department recognizes that transfer pricing is not an exact science and that the application of the most appropriate transfer pricing methodology may produce a range of results. The facts and circumstances of each case will be relied upon to determine either that range or a point in that range that gives the most reliable estimation of an arm’s length price.
 While the Act itself does not mandate methods or impose an explicit hierarchy of methods to be used in order to meet the requirements of the arm’s length principle, it is clear that the OECD Guidelines state that the traditional transaction methods are preferable to the transactional profit methods, a point of view that is endorsed by the Department.
 The most appropriate method in a given set of circumstances will be the one that provides the highest degree of comparability between transactions. Once a taxpayer establishes comparability at a particular level within the hierarchy of methods, the taxpayer is not required to consider or apply a lower-ranking method. On the other hand, if the taxpayer cannot establish comparability at any level with one method, the other methods should be considered in order to determine the most appropriate method.
 What is also clear from the OECD Transfer Pricing Guidelines is the predominance of and preference for the application of the CUP method followed by the other traditional transaction methods over the less preferred transactional profit methods. It is important to emphasize that, in the absence of useful evidence of an uncontrolled transaction, it will be necessary to use another method. However, the OECD Guidelines suggest that it is not appropriate to use other methods just because data concerning uncontrolled transactions is difficult to obtain, and admonish that every effort should be made to make the necessary adjustments to the data so that it may be used in the CUP method. As the best method is CUP, the OECD suggests that the parties should try to locate comparable uncontrolled transactions where possible.
 The CUP method compares the price charged in a controlled transaction (i.e., between non-arm’s length parties) with a price charged in a comparable uncontrolled transaction (between arm’s length parties) either by comparing transactions between one of the non-arm’s length parties and an arm’s length party (“internal comparables”) or transactions between two arm’s length parties (“external comparables”).
 The CUP method provides the highest degree of comparability of all methods because it focuses directly on the price of a transaction and requires both functional and product comparability, but other traditional transaction methods may have to be used where there is not enough quality information available with respect to uncontrolled transactions or where it is not possible to reliably qualify the differences between controlled and uncontrolled transactions. The same rationale applies to the provision of intra-group services, which clearly exist in this case.
 In paragraph 64 of Information Circular 87-2R, the Department reiterates that the CUP method provides the best evidence of an arm’s length price and further explains that a CUP may arise where “the taxpayer or another member of the group sells the particular product [or services], in similar quantities and under similar terms to arm’s length parties in similar markets (an internal comparable).
 As stated, the transactional profit methods are generally only to be considered for use as a last resort, where no quality data is available to permit the application of the traditional transaction methods or where material differences in the data prohibit adjustments.
 Where there are doubts about the reliability of the results produced by a particular method the taxpayer should confirm those results by applying another method, the rationale for this being, that recommended methods should have some consistency in their results.
 This “testing” of methodologies is only required, in my view, if there are reasonable doubts about the reliability of the result. In the General Electric case, a failure to test was, appropriately, not fatal.
 In the situation here, we have the Appellant relying on the CUP method, while the Respondent relies on the TNMM, the method generally considered the least reliable and the lowest on the totem pole or in the hierarchy.
The Expert Opinions
 Each side had an expert clearly qualified to give expert opinion on transfer pricing issues and it is safe to say, having considered all relevant material, that the experts, namely Mr. Wall on behalf of the Appellant and Dr. Wright on behalf of the Respondent, strongly disagreed with each other not only in their conclusions but on the pricing methodology to be used as well.
 Each expert acknowledged the pricing methodologies available for consideration under Canadian transfer pricing rules, pursuant to IC 87-2R, and discussed the natural hierarchy of such methods and their application, all of which I have made reference to in my discussion of transfer pricing law above. It should be noted that Mr. Wall’s approach was consistent with the natural hierarchy cited by the Minister, starting with the application of the CUP method as the most reliable method, down to the transactional net margin method as the least reliable, while Dr. Wright, in effect, took the position that the CUP method and other methodologies higher in the hierarchy did not apply and went straight down the list to utilize the TNM method.
 Each expert’s testimony paid homage to the Department’s position, set out in IC 87-2R, that a number of factors influence the degree of comparability of transactions, such as the characteristics of the product or service, the functions performed (taking into account the risks assumed and the intangibles), the terms and conditions of the contract, the economic circumstances, and the business strategies of the parties; and of course the two experts came to opposite conclusions.
 Mr. Wall reached the following conclusions in his report:
1. The internal CUP provides the most appropriate and reliable transfer pricing method. On this basis, the actual price charged by APCI to AP Circuits is arm’s length.
2. The external CUP is relevant and, after further analysis, is shown to be reliable. This information is best used as additional support for the internal CUP.
3. Ms. Bruce’s explanation and analysis for using the cost plus method in Fiscal 1999, and Mr. Laudano’s [the CRA’s internal economist] explanation for rejecting the cost plus method, appears insufficient or lacking. It is unclear why Ms. Bruce [the CRA’s auditor] switched to the TNMM method for Fiscal 2000 and 2001.
4. Part I of the TNMM analysis in this report concludes that AP Circuits (not APCI) might be the “least complex”. Part II shows what conclusion the Respondent might have reached if they had chosen AP Circuits (not APCI) as the tested party. For example, it appears AP Circuits earned an operating margin that falls within the arm’s length range defined by 12 companies that manufacture bare printed circuit boards. This is done for illustrative purposes only, and should not be read as a conclusion on, or support for, the TNMM as the preferred method.
 The Respondent, in cross-examination, made it clear she disagreed with these conclusions and generally attacked both Mr. Wall’s analysis supporting his conclusions, as well as the appropriateness of the fundamental assumptions made by Mr. Wall. Specifically, the Respondent argued that in applying the CUP analysis in particular, Mr. Wall inappropriately separated the set-up base fee from the square inch fee or bonus, as it was described in the Functional Analysis, it being the Respondent’s position that both were components of the set-up fee. It should be noted that Dr. Wright also made the same assumption in her analysis rejecting the CUP method of pricing and assumed as well that Appellant could and did perform part of the set-up function from time to time. I have already discussed above these disagreements regarding the factual assumptions and reiterate that I agree with the Appellant’s position on these matters. Accordingly, it stands to reason that the Respondent’s expert was in error in accepting such assumptions for the purpose of her report, and her report is highly suspect as a result.
 As regards his conclusion No. 1, Mr. Wall determined, as there was evidence that the prices the Appellant charged to its arm’s length customers were the same prices as APCI charged it, that these were relevant transactions and hence were an internal comparable. The Respondent argued that the parties had different levels of market or economic comparability, in the sense that the Appellant was a manufacturer and APCI provided solely set-up services, and so a relevant transaction would only be one between APCI as seller and an arm’s length manufacturer or between the Appellant as purchaser and an arm’s length supplier of set-up services. Dr. Wright, who also agreed with the hierarchy approach, stated that the CUP method is best if it can be applied, but found it not to be applicable here, saying there were no internal CUPs because APCI did not sell set-up sources to an arm’s length party, nor did the Appellant buy set-up services from an arm’s length party during the period at issue.
 With respect, this position of to the Respondent has several weaknesses. Firstly, the Appellant and APCI did provide the same services, in my view, since the Appellant was not just manufacturing circuit boards but, in effect, was also providing set-up services to its arm’s length customers, as advertised on its website. Notwithstanding that it may have outsourced those services, it was still selling such services to its arm’s length customers on the open market, billed set‑up services as a separate item to those customers, and was accountable to those customers for the quality of the services provided.
 Secondly, since APCI was providing those same services to the same arm’s length customers ‑ albeit the North American customers of the Appellant ‑ at the same price, one may conclude that they operated in exactly the same economic market. I note that the Respondent’s counsel suggested to Mr. Wall that the transfer pricing rules must exist in the context of what make business sense. In my view, it makes perfect business sense to treat services provided to a client through outsourcing in the same market as if they were supplied directly to the customer. What is necessary is to consider internal sales, whether by one member of a group, or by the group collectively, selling to arm’s length customers in the same economic market, which, in the present case, is the North American market. To find otherwise would mean accepting that there can never be internal CUPs in a situation where a related company provides services to, or manufactures only for, a member within its group.
 Moreover, the Department clearly accepts this view in its definition of an internal CUP, found in paragraph 64 of IC 87-2R:
64 The CUP method provides the best evidence of an arm’s length price. A CUP may arise where:
· the taxpayer or another member of the group sells the particular product, in similar quantities and under similar terms to arm’s length parties in similar markets (an internal comparable);
 The Respondent’s expert witness is clearly wrong when she states that in order to find an internal CUP the sale must be one made by APCI to an arm’s length party. The wording of the paragraph above clearly states that the sale can be by the taxpayer or by another member of the group. Accordingly, if evaluated from the perspective of APCI, the other member of the group is the Appellant, which clearly sells set-up services to arm’s length customers. If viewed from the perspective of the Appellant, it is itself the “taxpayer” that clearly sells the same service to arm’s length customers.
 Thirdly, the Respondent’s concern, that the Appellant and APCI are not in similar economic circumstances because they operate at different levels in the market was also couched in the assumption that APCI was a service provider providing the set-up function and that the Appellant was in effect a distributor of those services; which, of course, is an accurate description of the arrangement. Dr. Wright, in noting that the same prices were charged at both levels, indicated that the question to be asked is why the Appellant would give up the profits at the set-up level, which she indicated were approximately 40% at gross margin, and not charge a higher price on distribution. To use the Federal Court of Appeal’s test as articulated in Glaxo, the question can be rephrased to simply ask: Why would a reasonable businessman in the Appellant’s shoes pay the same price?
 The answer to those questions, from the evidence, is that the Bambers were knowledgeable about manufacturing but had little knowledge of or affinity for software use and little know-how with respect thereto and relied heavily on Mr. McMuldroch. Contracting out the set-up function, a manual process when the Bambers started out in business but which developed into a more high‑tech process involving custom software development and use, required them to seek out the services of people like Mr. McMuldroch who allowed them to focus on their core business of manufacturing while still controlling the client by providing the client with both the service of set-up and the product of manufacturing. The evidence showed that the Appellant’s sales rose dramatically as a result of this arrangement. In addition, the Bambers were, in effect, now in an arrangement where they kept 100% of the profits from manufacturing and still got two-thirds of the profits from APCI, as per their deal. It is hard to imagine in light of this, why a reasonable businessman would not agree to the same price being charged at the different market levels. Frankly, based on the gross profit margin of the Appellant before the Barbados move, which Ms. Wright calculated at 45%, it would appear the Bambers received the benefit of 75% of that amount (due to their ownership percentage) or 33% of the gross margins, while under the new arrangement they received approximately 32% (all 5% of the gross manufacturing profit and 66.67% (i.e., two-thirds) of the 40% service margins); however, with the much larger manufacturing sales volumes and hence increased set-up services as well, the overall pie that was shared was bigger and the Bambers were better off. I am satisfied that the above reasons alone would satisfy Dr. Wright’s inquiry, but I can add that not losing the valuable services of Mr. McMuldroch was an equally important factor.
 Fourthly, the Respondent’s view that the Appellant’s use of the internal CUP (which involves “unbundling” one transaction or separating the manufacturing and set-up components) is a fundamental breach of the principle against unbundling price components of one transaction seemed to be based on an erroneous interpretation of what the transactions were and seems to ignore the Department and the OECD’s views. The Department, as stated above, has no difficulty in unbundling transactions and, in fact, clearly expresses a preference for separating components of a transaction where quality data is available to set the price for separate transactions. In fact, only if such data is not available should the transactions be bundled in the first place according to the information circular. In the case at bar, the Appellant sells both the manufactured products (circuit boards) and the set-up service. It prices the two components separately, as evidenced by its invoices to its arm’s length customers, and as advertised on its website, where the method for pricing the two components is clearly set out. It stands to reason that, if the taxpayer is selling an apple and an orange to an arm’s length party, it should be an easier task to test the prices for the components separately. Frankly, the Respondent seems to ignore both her own views in the information circular as well as its own suggestion that what must be done must make business sense, as per Dr. Wright’s comment.
 Finally, even if the Respondent could demonstrate that the bundling of the separate components was a marketing tool leading to indivisible blended pricing or that the different market levels of supplier and distributor should, on the facts, justify a different price, then it seems to me that these would be the very differences in the comparability factors that the parties are required to make adjustment for when possible under the accepted process. Likewise, the differences that, the Respondent has claimed, exist in the comparability factors of risk assumed and of ownership of intangibles, as well as the functional differences whose existence she initially claimed, above, are all differences that should have been considered and argued in the context of making adjustments to the internal CUP in this case. More will be said of these other differences in later discussion, as they are more relevant in determining which party is to be the tested party when using the TNMM than they are in the CUP method, which focuses on price. The problem with the Respondent’s position on the use of the CUP is that she has made no attempt to analyze either the applicability of the CUP or differences in the factors for determining what, if any, adjustments to the prices used could be justified, and this constitutes a fundamental abdication of her duties under the transfer pricing rules.
 After analyzing the comparability factors, a copy of which analysis is attached hereto as Appendix “A”, Mr. Wall concluded as follows:
The above information and analysis shows, in accordance with the definition for an internal CUP, that AP Circuits sold the same base set up as a component of the bare circuit board, in the same quantities and under the same terms to arm’s length parties in the same markets. It is therefore reasonable for AP Circuits to have relied on this transaction for pricing the base set up services that it purchased from APCI, and to conclude that the amount paid was arm’s length.
Although the definition for an internal CUP, provided in paragraph 64 of the IC 87-2R refers to the comparison of features being similar, it is an important distinction that the above comparison shows the transactions are the same in all factors, which is a higher level of comparability.
 I think it is important to stress the fact that Mr. Wall found the same and not just similar comparisons. Not only did he review the invoices of the Appellant to its arm’s length customers, but he also reviewed the invoices of APCI to the Appellant, which contained matching terms and conditions in the description of both set-up fees and extras, and also in the payment terms. In addition, Mr. Wall analyzed the terms and conditions found on the Appellant’s website at the relevant time, which were also the same. The website provides further evidence that the public had the benefit of published rates. Mr. Wall also concluded that, since the Appellant was APCI’s only customer, the volume of its services provided (the number of set-ups) was identical to the volume sold by the Appellant to its customers. I agree with his analysis.
 The Respondent, of course, did not rely on this method and simply failed to make any effort to consider it, notwithstanding the clear hierarchy of methodologies and the preference for the application of the CUP method, particularly with respect to the provision of services. It is clear the CRA was aware that these internal comparisons were of relevant transactions but made no further inquiry or analysis to test them. As Mr. Wall has pointed out, during on‑site interviews at the Appellant’s offices on November 15, 2001, Mr. Sell of the CRA stated the following:
I then point out that APCI was paid the same as what APC [AP Circuits] charged its customers and that APCI was provided with this information.
 During the examination for discovery of Ms. Bruce of the CRA on October 15, 2009, she confirmed she never went through the invoices to determine whether they were the same, notwithstanding that she agreed that the proper method to follow would have been to get out the invoices and find out whether the facts were true – i.e., whether the same prices were charged.
 In his report, Mr. Laudano, the CRA’s economist, went so far as to state:
From the information provided there is no evidence of a CUP, whether internal nor [sic] external.
without further explanation as to why the transaction did not qualify as an internal CUP.
 It appears, to use a colloquialism, that CRA simply ignored the transactions and “dropped the ball”. From the evidence, it was clear the Appellant provided volumes of information and data sufficient to identify all invoices between the Appellant and its customers during the taxation years in dispute, as well as in years before and after those years to show consistency, and that it also provided the memorandums and invoices between APCI and the Appellant. The Respondent simply chose not to analyze them on her own.
 The published rates were readily available on the website of the Appellant and this alone could have created the basis for starting to test the terms and conditions in the annual contracts that were in the CRA’s possession. Again, the Respondent failed to consider this information.
 To conclude on this issue, in my view, the transfer pricing rules, policies and processes enunciated in the OECD Guidelines and the Department’s IC 87‑2R clearly set out the overwhelming preference for analysis of the highest methodologies first, before consideration of lower methods, and a preference for adjustments to prices where comparability factors justify a price differential. The evidence in this case is clearly that internal CUPs existed and were inappropriately ignored by the Respondent. I find the Appellant’s use of the internal CUP both acceptable and clearly within the preferred mechanisms of the rules and that the Respondent has not justified any adjustments to the Appellant’s finding that the prices paid to APCI for the set-up services based on the fixed rates charged, were arm’s length prices. Accordingly, 77.99% of all amounts paid by the Appellant to APCI were for set-up services only and were acceptable as prices that would have been paid by an arm’s length party. The balance of the fees paid by the Appellant to APCI, that is, the other 22.01%, will be discussed later.
External CUP to Test Internal CUP
 As the Federal Court of Appeal found in General Electric, there is no requirement that a method, if used, be tested, and, in my view, the transfer pricing rules make it clear that once the highest methodology applies and is used, that there is no requirement for an Appellant to go further. However, in line with the rationale in the OECD Guidelines that different methods should yield consistent results and hence enhance the initial conclusion, I note that Mr. Wall also undertook an analysis using the external CUP method ‑ which compares prices between arm’s length parties ‑ including consideration of the external comparable, namely Star Electronics, Corp., identified by the Appellant in its Functional Analysis submitted to the CRA.
 With respect to Star Electronics, Corp., Mr. Wall conducted an analysis of the comparability factors, not just from information contained in the letter provided by that company but also from a detailed Web search of Star Electronics, Corp., the results of which are set out in Appendix “B” annexed hereto, and he came to the following conclusions on page 33 of his expert report:
The above information and analysis shows, in accordance with the definition for an external CUP, that AP Circuits sold the same product as Star Electronics, in similar quantities and in similar markets.
The letter provided by Star Electronics (see Tab 18) indicates they charged a price ranging from US$100.00 to US$125.00 in 1999, which is significantly higher than the price of US$46.00 charged by AP Circuits to its customers in 1999 (Tab 17). Unfortunately, there is no additional information that supports the information provided by Star Electronics, which diminishes its reliability. However, there is no information that contradicts it, which suggests that it can still be used.
In general, unless there are specific reasons to exclude third party data, but not sufficient reason to rely on it as the primary method, the information is often used to support the arm’s length price or range defined by another, more reliable, transfer pricing method.
 The Respondent strongly disagreed with Mr. Wall’s conclusion that, since there was no information that contradicted the reliability of the Star Electronics, Corp. information, it should be considered reliable, and on this point I must agree with the Respondent. Reasonable steps must be taken to test the reliability of the suggested external CUP by both the party asserting its applicability and the party disputing it. In this case, again, the Respondent and her expert witness dismiss the possibility of utilizing the Star Electronics, Corp. figures as an external CUP, and they do so on the sole basis that there was not enough information in the Star Electronics, Corp. letter, without making any attempt to ascertain the veracity or weaknesses of the information. The Appellant’s expert witness, on the other hand, conducted a more thorough investigation of Star Electronics, Corp. and other potential competitors before arriving at his conclusions.
 I find it difficult to accept the Respondent’s rationale for not bothering to follow up on the comparable brought to her attention and relied upon by the Appellant. Again, the CRA simply ignored Star Electronics, Corp. as a potential comparable, notwithstanding the acknowledgment of Ms. Bruce, on her cross‑examination, that she was aware the Appellant was representing Star Electronics, Corp. “to be a comparable”.
 Oddly enough, Mr. Laudano concluded on page 2 of his report, that:
It is CRA’s position that the methodology provided by the taxpayer is inappropriate in determining the arm’s length price for the controlled transaction between APCI and APCL.
 One of the two reasons is that there was no indication as to what Star Electronics, Corp. did, or what their set-up fees were for, or how their process worked as compared to the Appellant and APCI, but the Respondent clearly made no effort to find this information out; that is particularly noteworthy when such information may have been readily available not only on the company’s website, as Mr. Wall discovered, but also directly from Star Electronics, Corp. which, in the letter referred to in the Functional Analysis, clearly invited further questions or the expression of any concerns. The contact’s name and the company’s address and telephone number were clearly visible for the CRA’s use. I am in agreement with Mr. Wall that the information found on the website, as it existed at all relevant times, described Star Electronics’ processes, product, use of data, and set-up in sufficient detail not only to allow a reasonable comparison to be conducted but to at least invite a preliminary analysis and search for any differences that could be used in making adjustments.
 The second reason provided by Mr. Laudano for the inappropriateness of the Appellant’s suggested external CUP was that he basically did not consider a quoted price to have the strength of a negotiated price, and hence ignored it. Interestingly enough, the letter actually states that “[the cost to prepare panelized data and drill route program in 1999 was approximately] $100.00 to $125.00.” Since the letter was dated July 11, 2005, a more reasonable interpretation would have been that he was referring to the actual price charged to their customers in 1999, as Mr. Wall suggested. Moreover, considering how much more Star Electronics, Corp. charged, i.e., from $100 to $125 versus $46 as the base set-up fee charged by the Appellant, it is certainly reasonable to conclude that there was over a 200% difference that the Appellant would have had as leeway in presuming its prices were competitive relative to the so-called “arm’s length quote”, and that should have at least garnered enough interest to warrant further inquiry.
 To be clear, counsel for the Respondent, in his cross-examination of Mr. Wall, also attacked the appropriateness of the external CUP method on the basis that Star Electronics, Corp., both as a supplier of set-up functions and as a manufacturer of circuit boards, could not be compared to APCI, which performed only set-up functions, and by necessary implication could not be compared to the Appellant, which handled the manufacturing function. The difficulty I have with this rationalization is, again, that the Appellant was in the business of selling set-up functions as well as manufacturing circuit boards for arm’s length customers. Both corporations operated in the North American market and, as indicated by Mr. Wall’s analysis, had similar comparables. Moreover, comparables are required to be similar and not necessarily the same under the transfer pricing rules. In any event, I still fail to understand the Respondent’s reluctance to accept that a party that only has one function could be compared with an arm’s length party that has several functions if these functions can be unbundled so that apples to apples are being compared, as this is the process recommended in paragraphs 36 to 43 of IC 87-2R referred to earlier.
 It should be noted that Mr. Wall’s analysis did not stop with Star Electronics, as he in fact also did searches for the 12 competitors the Appellant had provided to the CRA, and found that 8 of them had websites during the relevant period, and that 2 of those 8 had published their prices. He found that those two competitors had a base fee ranging from $50 to $75, depending on whether a single- or double-sided application was involved, which fee was obviously still higher than the $46 rate charged by the Appellant and APCI in the same year. While I agree that Mr. Wall confirmed one of the comparables as being at best a “loose” comparison, the other comparable was, in his opinion, sounder, and the Respondent made no attack on that comparable. That comparable alone would have given an external CUP of $50 for a single‑sided circuit board and $70 for a double-sided one. The point is that the transfer pricing rules require that an attempt be made to obtain comparables using the CUP method regardless whether it is difficult to do so and that every attempt be made to distinguish differences through adjustments to the price. This process was clearly not followed by the Respondent. Moreover, it is absolutely clear that there are many competitors in the market that could have been compared, including Western Index Manufacturing, whose principal, James Malcolm Godfrey, testified at trial that he charged in the neighbourhood of $120 for set-up services in years preceding the years in dispute, as evidenced by invoices submitted as evidence, and that due to the competitiveness of the market, set-up fees remained the same over the taxation years in question. While his testimony cannot be considered conclusive as to the going rates during the taxation years in question, it is evidence that there were many competitors in the market that could have been tested and his price was within the Star Electronics, Corp. range, adding further credibility to it as an external comparable.
 It is interesting to note that much of the information regarding these comparables was simply obtained by Mr. Wall using the internet, which information would have obviously been available for the CRA as well.
 What is also interesting is the great effort the CRA made using websites, to obtain and analyze information on competitors for the purpose of its preferred transactional net margin method (“TNMM”), competitors whose total sales ranged from $120 million to $964 million, and which had assets ranging from $332 million to $1.105 billion and 700 to 8,700 employees, instead of using as a comparable Star Electronics, Corp., which had total sales of $8 million and 70 employees, while APCI had total sales of $2.4 million and 5 to 8 employees during the same period. It seems obvious that Star Electronics, Corp. was a much closer comparable. I find it difficult to accept the rationale for the CRA not using more comparable entities for its own analysis, but even more difficult to accept that the CRA demonstrated that appropriate comparables were out there if one looked but failed to use any. That it would search for comparables for the lowest ranking method of establishing arm’s length prices and not for the highest method is perplexing to say the least.
 I must conclude, at this point, that the Appellant has satisfied me, using the internal CUP method tested by the external CUP method, that the prices it paid for set-up services were consistent with prices charged for set-up services between arm’s length parties. I intend to address the issue of the balance of the fees paid by the Appellant to APCI as well, but feel it is necessary to comment first on the other pricing methodologies referred to in this trial, particularly the Respondent’s use of the TNMM and my concerns regarding it.
 Dr. Wright came to the conclusion that the TNMM was the best method to apply because total cost information was available.
 Dr. Wright, of course, dismissed the use of the CUP method, both internal and external, for the reasons above discussed, but also dismissed the cost-plus method, which is based on a comparable mark-up on gross profits, on the sole basis that there was no information available from APCI’s financial statements to permit utilization of the cost-plus method. In other words, APCI did not separate its direct costs from its total operating costs so as to present a cost-of-goods-sold figure on its financial statements, which is the usual basis for using the cost-plus method. It is interesting to note that Mr. Laudano, the internal economist for the CRA, used that method for the 1999 taxation year, using a 5% mark-up without any explanation for that rate, and switched to the TNMM for the next two years in issue, also without explanation. Frankly, with the financial statements and books of account of APCI being available, it begs the question why such direct costs could not have been calculated; clearly, no further effort or attempt was made by the CRA or Dr. Wright to do so. Likewise, Dr. Wright made no reference to the resale price method or the profit split method, again without explanation.
 Dr. Wright agreed that once the TNMM is selected, the party to be tested should be the simplest one, and that one must look at the functions, risks and intangibles, referred to above, to determine which party is the simplest since the transfer pricing rules dictate that the simplest party should be the tested party using this methodology. This process is not disputed by Mr. Wall. Dr. Wright indicated that, since she considered the Appellant to have the greatest risks, as it owned, in her view, most of the intangibles, and performed all of the manufacturing function and part of the set-up function, managed the overall operation of the company, developed the market for the circuit boards, performed all marketing, selling and customer relations functions, and did invoicing and shipping, it was the more complex, and, by comparison, APCI the simpler party. Mr. Wall, on the other hand, concluded that APCI bore the greatest risk since it only had one customer, the Appellant, and owned the “valuable” and “unique” intangible assets, namely the software copyright and improvements, had the technical know-how regarding the use of the software programs, and performed the more complex set-up function and, accordingly, the Appellant was the simpler party and the one that should be tested. Needless to say, there is strong disagreement between the parties as to which has the simpler party and hence the one that should be tested. I do not intend to undertake a long analysis of the relative merits of both parties’ positions on this issue except to say that I have some serious concerns with some of Dr. Wright’s assumptions, as discussed below.
 Firstly, in her functional analysis Dr. Wright assumed that the set-up fee included the square inch fee paid annually and, further, that the Appellant continued to do part of the set-up as discussed earlier, which I found not to be supported by the evidence. Accordingly, her reliance on these two major factual assumptions found to be in error bring her analysis seriously into question.
 Secondly, Dr. Wright’s assumption that the software was being developed by Mr. McMuldroch directly, or under his supervision, while he was an employee of the Appellant before moving to Barbados and her reliance on the fact that APCI paid the Appellant nothing for any of its rights in such pre‑Barbados development and that the Appellant thus “gave” the intangible to APCI, in addition to her insistence that the payments made by the Appellant to APCI in order for APCI to continue to develop the software while in Barbados suggests that the Appellant paid for the development and hence was the owner, are all seriously flawed. The issue of the value of the software that may have been given is a valuation issue that the CRA apparently did not see the need to address. Paying for software development and maintenance does not automatically result in ownership of that intangible. Clearly, there is no question that, as Lavenir was paid for its software licences, ownership of such software still had not passed to the Appellant or APCI. In the absence of evidence of beneficial ownership of or actual title to the software, Dr. Wright assumes too much regarding the ownership of that intangible. In my view, ownership of the software improvements in question appears to have been with APCI not the Appellant. Moreover, I am in agreement with Mr. Wall that, although there is no question the Appellant owned the client list and manufacturing location, which I consider to be goodwill, and also as initially developed the cheaper, serial method of manufacturing circuit boards, which other competitors eventually adopted, these are not intangibles that are “unique” and “valuable” in the sense contemplated by the transfer pricing rules; but are, rather, routine intangibles that no longer distinguished the Appellant from its other competitors. At the very least I could have concluded that either APCI or both the Appellant and APCI owned or contributed valuable or unique intangibles and, accordingly, the TNMM method was not the preferred method to apply in such circumstances. In any case, paragraph 94 of IC 87-2R states that:
. . . the TNMM is less likely to produce reliable results where the tested party contributes to valuable or unique intangible assets.
 Thirdly, when one evaluates the market risks both sides were exposed to, there is no doubt both bore risks. However, Dr. Wright’s suggestion that, since the annual contracts did not address the question of who was liable for potentially poor set-up work or the warranty for such work, the Appellant took on that warranty risk is inconsistent with the evidence that adjustments were made each month for any work that had to be redone; it seems APCI’s payments were reduced as a result. In addition, the evidence was that the Appellant collected its fees in advance from its customers while APCI was paid 30 or more days later, so it would seem APCI may technically have been at greater credit risk due to delay in payment, although, practically speaking, when the party that makes the payment also benefits from two-thirds of the profit from it, it becomes rather hard to suggest that there was much risk of non-payment. On the other hand, I agree with Mr. Wall’s assertion that since APCI had only one customer, namely the Appellant, if APCI lost that customer, it would have been for all intents and purposes finished in business and hence bore the biggest market risk. This is what in fact happened. To suggest, as Dr. Wright did, that since APCI was handed a guaranteed market by the Appellant, resulting in high profitability from day one, seems to ignore the reality that having one’s eggs all in one basket means you can be out of business at the whim of that sole customer as well.
 It was also agreed that both sides had some foreign exchange currency risk since both were paid to some extent in U.S. dollars, and that both sides had investments in their operations and equipment. Hence, there must have been at least some investment risk on both sides as well.
 As further evidence to support her position that APCI was the simpler party and thus should be tested, Dr. Wright wrongfully assumed APCI did over 500 set‑ups per day, suggesting the set-up process was very simple if the relatively unskilled staff of APCI could do so many per day. The evidence, however, shows that APCI averaged only about 50 set-ups per day during the taxation years in dispute, only one-tenth of the number assumed by Dr. Wright. This admitted error, in my view, is clearly indicative of the further errors committed by Dr. Wright in interpreting the Functional Analysis of the Appellant and diminishes the credibility of her report.
 While, having regard to the evidence, I disagree with many of the assumptions made by Dr. Wright, it is also clear that Dr. Wright, as well as Mr. Wall, made subjective interpretations of the evidence presented to them. When, as in this case, subjective opinions are capable of yielding such substantially different results, it is clear that the TNMM lends itself to inconsistent and contrary conclusions, which emphasizes the reason that this method is considered the method of last resort.
 In fact, the analysis undertaken by both expert witnesses led to opposite results, since each tested a different party. A short comparison of the analysis of both experts sheds further light on the unreliability of the TNM method.
Dr. Wright’s TNMM Analysis
 In establishing the mark-up on cost for service providers, Dr. Wright tried to identify publicly traded companies that provided set-up as a part of their business activities, and, not finding any that only did set-ups (because she concluded most performed all functions of set-up and manufacturing), she moved on to consider “broadly comparable” businesses in order to benchmark mark-ups on total cost. To do so, she identified service companies that had similar employee profiles (degree of training, etc.) and similar risk profiles, and that utilized intangible property (primarily software) to render services. To establish a lower range of mark-ups, she looked at service providers whose employees had little training and formal education, and which used relatively low-value intangibles. To establish the upper range, she looked at service providers whose employees were more highly technically trained and used higher-value intangibles to provide value to their clients. Her analysis of several companies, generally described as business services companies, showed that the lower range over the 1997 to 2001 period averaged margins between 4.1% and 9.6%, or, if one knocks off the top and bottom 25%, one is left with an interquartile range of 2.0% to 9.9%. Accordingly, she concluded that a comparable range starts at the lower point of 2.0%. Using a similar method, based on four comparables generally described as contract programming companies, the interquartile upper limit was set at between 6.7% and 10.7%. Accordingly, while the range was set at between 2.0% and 10.7%, she utilized the most generous percentage of 10.7% to arrive at her figures, which, subject to the concessions made by the Respondent (who conceded the higher amounts where her internal economist found them to be appropriate for the 1999 taxation year), resulted in the transfer pricing adjustments now in dispute. In summary, she gave the Appellant the benefit of the highest mark-up and still concluded the prices paid by the Appellant to APCI were higher than the prices the Appellant would have paid to an arm’s length party.
Mr. Wall’s TNMM Analysis
 Mr. Wall, in fact, tested the TNMM using the Appellant as the test subject against twelve comparables. All were circuit board manufacturing competitors he culled from 20 possibilities using financial information, as well as information as to their processes available online. He concluded that the Appellant’s operating margins clearly fell within the range of margins established by those comparables for each of the taxation years in question, as shown in the chart on page 42 of Mr. Wall’s report, reproduced below:
 Accordingly, one must agree with his conclusion that, if these operating margins were within the acceptable range, then it seems reasonable to conclude that the prices paid to APCI must have been competitive transfer prices.
 Notwithstanding the strong argument, based on principle alone, that use of the TNMM by the Respondent was inappropriate, Mr. Wall also clearly demonstrated to the Court that one of the reasons the TNMM can be so unreliable is that it may yield wildly different results depending on which party is the tested party, as was the case here.
 What is also notable comparing the two analyses is that what Mr. Wall looked at as comparables were enterprises in the circuit board business while Dr. Wright took the much broader view that service businesses with employees performing more general services (but with ranges of skill and technical know‑how below or at the level of the APCI employees and also above their level) could be used to construct a range of margins. She compared general business services companies as the lower range with contract programming companies for the higher range. In doing so, she even eliminated from consideration the circuit board manufacturers used by Mr. Laudano, the CRA’s internal economist, in arriving at his assessments. Certainly Mr. Laudano’s chosen manufacturers were by and large much larger competitors than the ones used by Mr. Wall, but Dr. Wright’s were even larger entities, resulting in even less comparable entities being selected for comparison purposes.
 This analysis demonstrates the incredibly subjective aspects of TNMM, as not only are the assumptions made by each expert in applying the method different, but the market comparables go from direct competitors, i.e., printed circuit board manufacturers, to businesses in broadly comparable service industries, a far more general and indirect comparison. Dr. Wright seems to have ignored the value of having direct or substantially similar comparables in favour of an approach using broadly comparable factors, one which, in my view, leads to the inevitable conclusion that such an approach is the more unreliable of the two TNMM analyses.
The Fee for Other Services
 Having found that 77.99% of the total amounts paid by the Appellant to APCI were solely for set-up services and that those amounts represented arm’s length prices, it follows that that percentage of the transfer price adjustment must be reduced. The sole issue now is to deal with the remaining 22.01%, which, having regard to the annual contracts, I have found to be for the other services, described in Part 1 and Part 3 of the annual contracts, namely software development and maintenance and website development and maintenance.
 The Appellant took the position that the sums involved were minor and should be considered to have been for services provided at arm’s length prices because Mr. Wall’s TNMM analysis shows that when you consider the total fees paid to APCI ‑ i.e., the base set-up fees and these other fees together ‑ the Appellant was left with a profit in the range seen for other circuit board manufacturers to which Mr. Wall compared the Appellant.
 The Respondent, of course, treated these other fees, as did Dr. Wright, as being entirely for set-up, which I have found is not the case. It should be noted, however, that in argument the Respondent took the position that “some” of the fees were for set-up and, accordingly, by inference, some must have been for these other services. The problem, of course, is that the Respondent put forward no arguments as to what part, if any, was for these other services and whether the prices were considered arm’s length prices, other than to suggest that, since the Appellant reduced the square inch fees by $50,000 in 2000 due to Mr. McMuldroch’s return to Canada during that fiscal year (which effectively left no one to adequately service customers from Barbados for 7½ months of that fiscal year), this amount must have been the real fee for such services.
 The fact of the matter is that apart from describing these other services, there was no evidence led by either party as to how much time was spent on them. Aside from general evidence of such other services provided, there were no detailed invoices listing the actual components of such services or what the arm’s length price for such services could be.
 With respect to the Appellant’s position, it makes little sense to rely on the TNMM analysis of Mr. Wall regarding this matter, as it was a method the Appellant and Mr. Wall argued first and foremost was unreliable and not the method to use in this case. It would have been more appropriate to follow the process the Appellant argued was the right process by attempting to find CUP comparables and then proceeding down the list of methodologies to determine the proper one after due analysis of the comparability factors, and then validating that analysis by use of the next highest applicable method to test the results, if necessary. One thing that is clear from the evidence is that the Appellant utilized the services of a software programmer before, during, and after Mr. McMuldroch’s Barbados period, so comparable uncontrolled prices should not be hard to find. Likewise, the Respondent used information regarding computer programming services in establishing the higher range of margins in Dr. Wright’s TNMM analysis, which further suggests that software programming and website programming and software and website maintenance were services for which comparable uncontrolled prices are available.
 It can, of course, be argued that the Minister did assume that the Appellant overpaid APCI by the aggregate amounts of the transfer pricing adjustments the CRA made for the three taxation years in question, as stated in paragraph 17(dd) of the Respondent’s Reply, and that since the Appellant has satisfied the Court that at least 77.99% of the total amounts supposedly overpaid represented arm’s length prices, the Appellant has met the onus of rebutting that general assumption. However, such would not be an appropriate result in its context, in my mind, especially since the Appellant has argued that the transactions should be treated as being unbundled for greater reliability in the first place. In this context, the Appellant has the same onus in rebutting the general assumption that it overpaid APCI for each of the transactions ‑ namely the set-up services and the other services in question ‑ it seeks to unbundle.
 Frankly, were it not for the admission of the Respondent that some of the amounts in question were for set-up fees, and hence at least some, by necessary implications must have been for these other services, I would have been inclined to find that the Appellant has not met that onus.
 In evaluating the evidence, it is abundantly clear that only Mr. McMuldroch performed the services referred to in Parts 1 and 3 of the annual contracts. The other employees of APCI, including Brandi Lewinske, did not have the background, education or training to perform software development and maintenance and website development and maintenance, and it is clear the Functional Analysis of the Appellant supports this view. The first sentence of the first full paragraph found on page 3 of the Functional Analysis of Services reads as follows:
Dan McMuldroch’s expertise was in computer software, computer networks, software design and implementations, web design, and the setup function.
 The Functional Analysis refers to only three employees employed by the Appellant that were involved in the set-up function prior to Mr. McMuldroch’s departure from Canada to Barbados and indicates that they were Brandi Lewinske, Mr. McMuldroch, and the less experienced David Wu. There is no evidence that anyone, other than Mr. McMuldroch, had experience or training in performing the functions referred to in Parts 1 and 3 of the annual contracts and there is no evidence that anyone was hired or trained in Barbados who was capable of performing them along with Mr. McMuldroch. In fact, the evidence is that APCI had great difficulty in finding employees in Barbados with sufficient basic computer skills to train for the set-up function, let alone these other functions.
 The Functional Analysis confirms on page 4 that:
. . . Again, of the three shareholders, Mr. McMuldroch was the only one who had any expertise in this area.
This was discussed in the context of the performance of the CAD/CAM (software) services and the building, maintaining, and refining of the website.
 It is also clear from the evidence that Mr. McMuldroch returned to Canada from Barbados on June 12, 1999, so that there was, in effect, no one in Barbados that could perform the non-set-up function after that. No such services could have been performed thereafter by APCI pursuant to the annual contracts.
 With respect to the other services performed up to June 12, 1999, there is evidence of changes having been made to the website, as determined through Mr. Wall’s investigations, and evidence that the website had become fully interactive by late 1998. Moreover, there is also evidence that Mr. McMuldroch had developed the APCI main software to interact with the Lavenir software, as well as improved billing and data inputting systems, around or before that time, but there are no further specifics in that regard. What is clear from the evidence, in my mind, is that when the Barbados operation started in earnest in January 1997, APCI was able to perform around 50 set-ups per day, which was more or less the effective average during the taxation years in question. It therefore stands to reason that whatever other services needed to be in place to allow it to do so were at least substantially already in place at that time. Moreover, the fact that payment for such services was considered a bonus and was made only once annually suggests it was not tied to any specific volume of work performed or time spent. The mere fact that it is expressed as a rate per square inch of circuit board size, being the same basis as that on which the Appellant charged its customers for manufacturing the circuit boards, suggests that payment on this basis was not founded on an arm’s length principle, especially when one considers that the programmers hired by the Appellant on contract during the taxation years in question were paid at least monthly and either on the basis of time or on a fixed-price basis. Accordingly, while the Appellant has made a prima facie case that it performed some of the other services for APCI during the period up to mid-June 1999, it is not possible to quantify such services from the evidence, let alone value those services on the basis of the arm’s length principle of the transfer pricing rules.
 As summarized in paragraph 56 of these reasons, the total square inch fees ‑ or bonus as it was called ‑ paid and claimed by the Appellant for tax purposes was $322,625.13 for 1999, $344,821.92 for 2000, and $293,121.88 for 2001. Of course, there is no dispute that APCI had a January 31 year-end on which date those fees were paid by the Appellant. That is the reason the Appellant deducted them for its 1999, 2000, and 2001 taxation years respectively, even though the payments reflected services provided by APCI in the Appellant’s fiscal year commencing February 1 of the year before that fiscal year-end date.
 The only evidence suggesting any price for the other services is found in the 2000 summary of the total of $344,821.92 paid, which, as noted, was net of $50,000 to reflect the fact that in effect no one was in Barbados to perform those services for a 7½-month portion of APCI’s fiscal year; on an annualized basis, the amount would thus have been $80,000. Since the square inch fee was based on $0.2285 per square inch in both the 1999 and 2000 years, it stands to reason that the Respondent would agree that the value of such services performed for the full year in 1999 must have been $80,000.
 As I have said, because the Appellant has not met the onus of establishing that it did not overpay APCI for these other services, I cannot grant it credit for amounts of square inch fees paid any greater than the amounts implicitly conceded by the Respondent referred to above.
 Accordingly, the Appellant would be entitled to claim the payment of $80,000 for 1999 and $50,000 for 2000 (to reflect the fact that Mr. McMuldroch did not work 7½ of the 12 months, or 62.5% of the year), but of none of the fees paid for 2001, during which year Mr. McMuldroch was employed in Canada.
 In other words, there should be a transfer pricing adjustment of the difference between the fees claimed for the other services for the years in question and the fees I have found could be claimed above; namely an adjustment of $242,625.13 for 1999, of $344,821.92 for 2000, and of $293,121.88 for 2001.
 Accordingly, the appeal with respect to the transfer pricing adjustments is allowed, and the matter shall be referred back to the Minister on the basis that there shall only be the following transfer pricing adjustments added to the Appellant’s income:
1) for the 1999 taxation year - $242,625.13,
2) for the 2000 taxation year - $344.821.92, and
3) for the 2001 taxation year - $293,121.88,
for a total of $880,568.93.
 As per the agreement reached between the parties at the outset of this trial, the penalties on the above adjusted amounts shall be calculated in accordance with the provisions of subsection 247(3) of the Act.
 The parties shall have 14 days either to advise the Court that they have reached a settlement as to the issue of costs in this matter or to submit to the Court their written representations on costs.
Signed at Ottawa, Canada, this 29th day of April 2011.