Tax Court of Canada Judgments

 

 

Docket: 2008-2949(IT)G

 

BETWEEN:

MCKESSON CANADA CORPORATION,

Appellant,

and

 

HER MAJESTY THE QUEEN,

Respondent.

____________________________________________________________________

Appeal heard on common evidence with the appeal of

McKesson Canada Corporation 2008-3471(IT)G on October 17 to 20, 2011, October 25 to 28, 2011, October 31 to November 2, 2011,

November 15 to 18, 2011, November 29 to December 2, 2011,

December 12 to 15, 2011, January 16 to 20, 2012 and

January 31 to February 3, 2012 at Toronto, Ontario

 

Before: The Honourable Justice Patrick Boyle

 

Appearances:

Counsel for the Appellant:	Paul B. Schabas Ryder Gilliland Jeffrey Trossman Ilan Braude Kaley Pulfer
Counsel for the Respondent:	Guy Laperrière Janie Payette Sylvain Ouimet Chantal Roberge

____________________________________________________________________

 

JUDGMENT

          The appeal from the reassessment made under the Income Tax Act with respect to the Appellant’s 2003 taxation year is dismissed, with costs, in accordance with the attached Reasons for Judgment.

IT IS ORDERED THAT:

 

The parties are to file written submissions on costs within 30 days or such later date that the Court may agree to within that time. The written submissions are to include advising the Court whether a hearing on costs is requested.

 

The parties are to advise the Court within 30 days of their proposal for promptly addressing the proper identification for the Court of any confidential information in the Court record prior to public access to the record being reinstated. This may take the form of a Case Management Conference.

 

The parties are to advise the Court in writing within 3 days if they believe any confidential information is contained in the Court's Reasons for Judgment which should be addressed before being released to the public. Until then, the Reasons for Judgment are released to the parties and their counsel only and are not to be further disseminated by any party. 

 

 

       Signed at Edmonton, Alberta this 13^th day of December 2013.

 

 

"Patrick Boyle"

 

Docket: 2008-3471(IT)G

 

BETWEEN:

 

MCKESSON CANADA CORPORATION,

Appellant,

and

 

HER MAJESTY THE QUEEN,

Respondent.

____________________________________________________________________

Appeal heard on common evidence with the appeal of

McKesson Canada Corporation 2008-2949(IT)G on October 17 to 20, 2011, October 25 to 28, 2011, October 31 to November 2, 2011,

November 15 to 18, 2011, November 29 to December 2, 2011,

December 12 to 15, 2011, January 16 to 20, 2012 and

January 31 to February 3, 2012 at Toronto, Ontario

 

Before: The Honourable Justice Patrick Boyle

 

Appearances:

Counsel for the Appellant:	Paul B. Schabas Ryder Gilliland Jeffrey Trossman Ilan Braude Kaley Pulfer
Counsel for the Respondent:	Guy Laperrière Janie Payette Sylvain Ouimet Chantal Roberge

____________________________________________________________________

 

JUDGMENT

The appeal from the reassessment made under the Income Tax Act with respect to the Appellant’s 2003 taxation year is dismissed, with costs, in accordance with the attached Reasons for Judgment.

          IT IS ORDERED THAT:

The parties are to file written submissions on costs within 30 days or such later date that the Court may agree to within that time. The written submissions are to include advising the Court whether a hearing on costs is requested.

 

The parties are to advise the Court within 30 days of their proposal for promptly addressing the proper identification for the Court of any confidential information in the Court record prior to public access to the record being reinstated. This may take the form of a Case Management Conference.

 

The parties are to advise the Court in writing within 3 days if they believe any confidential information is contained in the Court's Reasons for Judgment which should be addressed before being released to the public. Until then, the Reasons for Judgment are released to the parties and their counsel only and are not to be further disseminated by any party. 

 

 

       Signed at Edmonton, Alberta this 13^th day of December 2013.

 

 

"Patrick Boyle"

 

 

 

Citation: 2013 TCC 404

Date: 20131213

Dockets: 2008-2949(IT)G

2008-3471(IT)G

BETWEEN:

MCKESSON CANADA CORPORATION,

Appellant,

and

 

HER MAJESTY THE QUEEN,

Respondent.

 

 

REASONS FOR JUDGMENT

 

 

TABLE OF CONTENTS

 

1.       Overview

2.       The Financing Transactions

a) The RSA Discount Formula

(i)      The Yield Rate

(ii)     The Loss Discount

(iii)    The Discount Spread

3.       The TDSI Opinions on Arm’s Length Terms & Conditions and Pricing

a) Eligibility Criteria

b) Termination Events

(i)      Delinquency Ratio Trigger

(ii)     Loss Ratio Trigger

 

c) Discount Rate

(i)      The Yield Rate

(ii)     The Loss Discount

1.       Designated Obligors

2.       Other Obligors

(iii)    The Discount Spread

1.       Servicing Discount

2.       Prompt Payment Dilutions Discount

3.       Accrued Rebate Dilutions Discount

4.       Interest Discount

4.       TDSI’s Supplemental Report on Servicing Fees

 

5.       The Law

(a)     GlaxoSmithKline

(b)     Reasonableness

(c)      Relevant Series of Transactions

(d)     Scope of Adjustments Permitted Under Paragraph 247(2)(c)

(e)      Factors that Exist Only because of the Non-Arm’s Length Relationship

 

(f)      The Rule in Browne v. Dunn and Opinions Within a Witness’ Expertise

 

(g)     The Court’s Analytical Approach to be Followed in this Case

 

6.       The Position of the Appellant

7.       The Position of the Respondent

8.       The Witnesses, the Expert Reports and the PwC Report

(a)  Mr. Brennan

(b)  Ms. Hooper and the TDSI Reports

(c)  The PricewaterhouseCoopers Report

(d)  The Frisch Expert Report

(e)  The Reifsnyder Expert Report

(i) Factors One and Two – Servicing Fees and Prompt Payment Discounts

       (ii)  Factor Three – Credit Risk

       (iii) Factor Four - $900MM Commitment to Finance

(f)  The Becker Expert Report

(g)  The Finard Expert Report

(h)  The Glucksman Expert Report

(i)   The Other Expert Reports

 

9.       The Appropriate Methodology

 

10.     Analysis of Transfer Pricing Issue

a) The Discount Rate

(i)      The Yield Rate

(ii)     The Loss Discount

(iii)    The Discount Spread

1.       Servicing Discount

2.       Prompt Payment Dilutions Discount

3.       Accrued Rebate Dilutions Discount

4.       Interest Discount

b) Summary of Court’s Estimate of Discount Ranges

 

11.     Conclusion on Transfer Pricing Adjustment

 

12.     Timeliness of Part XIII Assessment of McKesson Canada

a) The Issue

b) The Provisions of the Income Tax Act and the Treaty

c)  Position of the Parties

d)  The Interpretation of Treaties

e)  Analysis

f)  Conclusion re: Part XIII and the Treaty

 

13.     Dismissal of Appeals

 

 

Boyle J.

 

[1]             The principal appeal by McKesson Canada Corporation (“McKesson Canada”) concerns the amount of the transfer pricing adjustment made by the Canada Revenue Agency (“CRA”) to its income under paragraphs 247(2)(a) and (c) of the Income Tax Act[1] (the “Act”) in respect of financial transactions involving McKesson Canada and several of its non-arm’s length and related non-Canadian affiliates during its 2003 taxation year. The related appeal involves the secondary issue of McKesson Canada’s liability under the Act for its failure to withhold and remit to CRA an amount equal to the Part XIII non-resident withholding tax from the disallowed amounts paid by it to its non-resident parent.

 

[2]             As described in greater detail below, in 2002 it was decided by the McKesson Group that McKesson Canada would sell the receivables owing to it from its customers to a related non-resident McKesson Group entity at a discount. A facility was put in place pursuant to which the receivables would be transferred by McKesson Canada daily at a discount from the face amount of each transferred receivable.

 

1.       Overview

 

[3]             McKesson Canada is the principal Canadian operating company in the McKesson group of companies owned by the U.S. multinational McKesson Corporation (“McKesson U.S.”). McKesson U.S. is a United States public company and is the 15^th ranking largest public company in the Fortune 20 list of companies. Its annual revenues are an excess of US$100 billion. It is the largest U.S. health care company. It has over 32,000 employees worldwide. It has been said that McKesson U.S. is the biggest company no one has ever heard of.

 

[4]             Worldwide, and in Canada, the core business of the McKesson group of companies (“McKesson Group”) is the wholesale distribution of over-the-counter and prescription pharmaceutical medicine products. This accounts for about 97% of its revenues. Its other related business is that of hospital software technology.

 

[5]             The McKesson Group’s wholesale pharmaceutical business has an impressive market share. The McKesson Group delivers one-third of all medicare to the public in the US. In the years in question, McKesson Canada had about one-third of the Canadian market. It distributes the products of a large range of pharmaceutical companies, and sells to drug store chains large and small, to large grocery store and department store chains that have pharmacies and/or sell over-the-counter medicinal products, to independent pharmacies, to hospitals, and to long-term care institutions.

 

[6]             In the year the receivables facility was put in place, 2002, McKesson Canada had sales of $3 billion, profits of $40 million, 2,400 employees and the largest share of the Canadian market. Its Canadian customers included a number of Canada’s largest retailer grocers and drug store chains. It had credit facilities available to it in the hundreds of millions of dollar range with major financial institutions. Its public ultimate parent, McKesson U.S., had a solid investment grade credit rating and the interest rates on the available lines reflected that.[2] The McKesson Group had a very considerable cash surplus built up in its Irish affiliate.

 

[7]             At that time, and in the years leading up to it, McKesson Canada had its own successful and sizeable credit department which managed its credit and collection policies and practices. Credit and collections results were trending favourably. McKesson Canada’s receivables were managed with considerable success, having a roughly 30 day payment average,[3] and a 0.043% bad debt experience[4] with its customers overall. That is, 99.96% of its receivables proved to be good and collected when managed by McKesson Canada’s credit department applying McKesson Canada’s credit and receivables collection policies. This was very important to McKesson Canada’s success given that the wholesale drug business was low-margin – in the range of 2% - on high volumes.

 

[8]             There was no evidence that there was any pending imminent or future change expected, anticipated or planned for in the make-up, nature or quality of McKesson Canada’s receivables or customers, although there was always the future risk of unforeseen adverse change.

 

[9]             At that time, McKesson Canada had no identified business need for a cash infusion or borrowing, nor did McKesson Group need McKesson Canada to raise funds for another member of the group. There was a so-called double-dip Nova Scotia Unlimited Liability Company or ULC financing which was coming to maturity and would need to be recapitalized in some fashion; this was for a fraction of the amount of the new receivables facility. McKesson Canada did not approach its traditional lenders or conventional financial institutions (nor anyone else) before entering into its own non-arm’s length receivables facility and related transactions. The McKesson Group had previously put in place a tax-effective international corporate structure and inter-group transactions that allowed it to amass very large amounts of cash in Ireland. The non-Canadian members of the McKesson Group were able to use this money to finance all of the purchases of McKesson Canada’s receivables under the facility.

 

[10]        The non-Canadian McKesson Group company that purchased the receivables had the right to put non-performing receivables back to McKesson Canada for a price equal to 75% of the face amount, later readjusted to the amount actually collected on it by McKesson Canada. The purchaser did not otherwise have recourse to McKesson Canada for unpaid purchased receivables.

 

[11]        The non-Canadian McKesson Group entity that purchased the receivables borrowed all the money needed from another non-Canadian McKesson Group entity. The borrower’s obligations to the lender under the loan was fully guaranteed by yet another non-Canadian McKesson Group entity, which also indemnified the borrower for any shortfall between what the borrower received from McKesson Canada’s receivables and what it needed to pay on the loan.

[12]        As described below, the non-resident affiliate also paid McKesson Canada to continue to have McKesson Canada’s credit and collections department manage the receivables applying McKesson Canada’s credit and receivables collection policies and practices. Under the agreements these policies and practices could not be changed without consent. Similarly, McKesson Canada could only continue to grant other discounts or rebates in the ordinary cause of its business and in accordance with its usual practices when the facility was entered into.

 

[13]        Most of the proceeds of the initial $460,000,000 receivables sale were returned by McKesson Canada to its non-resident shareholder affiliate, a portion was loaned for a period to another Canadian corporation to permit its tax losses to be used, and about 1% of the proceeds were used by McKesson Canada for its general corporate purposes.

 

[14]        The CRA has challenged these related party transactions for McKesson Canada’s 2003 taxation year on the basis that the amounts paid to the non-Canadian McKesson entity pursuant to the receivables purchase transactions differ from those that would have been paid between arm’s length persons transacting on arm’s length terms and conditions. The discount upon the purchase of the receivables in accordance with the revolving facility was a 2.206% discount from the face amount. While this discount rate and the overall transactions between the parties are considered in greater detail below, this discount rate for receivables that on average were expected to be paid within about 30 days can be restated as an annual financing cost payable by McKesson Canada for its rights under the facility in the range of 27% per annum.[5]

 

[15]        A direct result of these discounts was that McKesson Canada ceased to be profitable for its 2003 taxation year and reported a tax loss in the year in issue in this appeal. McKesson Canada’s profits in later years were similarly significantly reduced.

 

[16]        The taxation year of McKesson Canada under appeal ending March 29, 2003 was a short taxation year of approximately three and a half months, having started upon its amalgamation as part of a Canadian restructuring of the McKesson Group’s Canadian interests. Its taxation and financial year ends on the last Saturday in March of each year. Its financial year is divided into a 13 four week Accounting Periods. CRA’s 2003 transfer pricing adjustment was approximately $26,610,000, reflecting a 1.013% discount for the purchased receivables.[6] No transfer pricing penalty was assessed.

 

[17]        The receivables facility was a five-year revolving facility. As detailed below, the purchaser had several rights to terminate the agreement in the event of any anticipated deterioration in the quality of receivables generated in McKesson Canada’s business.[7]

 

[18]        As discussed further below, the predominant purpose of McKesson Canada entering into the transactions was the reduction of its Canadian tax on its profits. None of the raising or freeing up of capital, reducing credit risk from its customers, nor improving its balance sheet was McKesson Canada’s predominant purpose; they were results of the transactions.

 

[19]        This trial was a very lengthy and hard fought 32 day trial heard over a period of five months from October 2011 to February 2012. Formidable groups of lawyers represented each of the Appellant and the Crown. The Court heard from two material witnesses and five expert witnesses. Reams and reams of documentation were entered into evidence, including further expert reports whose authors did not testify. After oral argument both parties made further written submissions and further responding submissions. Following the Supreme Court of Canada’s decision in Canada v. GlaxoSmithKline Inc., 2012 SCC 52, [2012] 3 S.C.R. 3 in October 2012, both parties made further written submissions.[8]

 

2.       The Financing Transactions

 

[20]        McKesson Canada and its Luxembourg immediate parent company (“MIH”) entered into a Receivables Sales Agreement (the “RSA”) and a Servicing Agreement effective December 16, 2002.

[21]        Under the RSA, MIH agreed to purchase all of McKesson Canada’s eligible receivables as at that date (about $460,000,000) and committed to purchase all eligible receivables daily as they arose for the next five years unless earlier terminated, and subject to a $900,000,000 cap.

 

[22]        Eligible receivables were generally trade receivables owing by an arm’s length customer who was not in default on other receivables and whose receivables would not, except in specific circumstances described below, represent in the aggregate more than 2% of the then outstanding receivables pool. The 2% concentration limit on eligibility did not apply to McKesson Canada’s handful of its largest named customers who accounted for about one-third of sales and whose receivables each already exceeded 2% of McKesson Canada’s outstanding receivables pool (the “Designated Obligors”). All hospitals were also defined to be Designated Obligors. The RSA has specific provisions, including discount rate calculation considerations, applicable to these Designated Obligors. The RSA contemplated the 2% concentration limit being waived or additions being made to the list of Designated Obligors with MIH’s consent.

 

[23]        The RSA provided that if a termination event occurred, MIH could direct McKesson Canada to advise its customers of the sale. In accordance with the related Servicing Agreement, McKesson Canada would continue to service and collect the receivables in accordance with its credit and collection policies and practices which were not to be changed without the consent of MIH.

 

[24]        The Servicing Agreement provided that McKesson Canada was to be the initial servicer but could be replaced upon the occurrence of a termination event under the RSA. MIH agreed to pay a fixed annual fee of $9,600,000 to the servicer to service the outstanding transferred receivables regardless of the amount outstanding.[9]

 

[25]        McKesson Canada did not warrant or guarantee the collectibility of the receivables or any portion thereof. MIH had the right to put a defaulted receivable back to McKesson Canada for an amount equal to the lesser of (i) 75% of its face amount, and (ii) the amount ultimately collected on it. When exercised, McKesson Canada was to pay the 75% amount to MIH and any ultimate downward adjustment was to be made subsequently.

[26]        MIH could terminate its obligations to purchase any further McKesson Canada receivables upon the occurrence of certain defined termination events, generally designed to identify or anticipate deteriorating creditworthiness of McKesson Canada or its pool of customers generating the receivables. These events included financial defaults of McKesson Canada or its affiliates, increases in the delinquency ratio or loss ratio of the receivables beyond specific thresholds, a downgrade in the credit rating of McKesson U.S., McKesson Canada’s name being changed to drop the word McKesson, McKesson Canada ceasing to be controlled by McKesson U.S., McKesson U.S. ceasing to guarantee McKesson Canada’s bank and commercial paper lenders, and any event occurring which materially adversely affected the enforceability or collectibility of the receivables or MIH’s rights under the agreements.[10] It can be noted that the termination events were not limited to things in McKesson Canada’s control, and included events in the control of its direct and indirect shareholders/parent corporations.

 

[27]        McKesson Canada continued to collect the receivables in the ordinary course. While ownership of the receivables was transferred daily, settlement (i.e. payment by the purchaser MIH) was more or less monthly.[11] McKesson Canada was not required under the RSA or the Servicing Agreement to segregate the funds collected on MIH’s behalf as they came in, unless MIH required it following a termination event. Each month the amount collected on the receivables for MIH’s benefit would be used first to pay McKesson Canada for newly generated receivables and any balance would be remitted to MIH. If there was a shortfall because newly originated receivables exceeded amounts collected, MIH was required to put McKesson Canada in funds.[12]

 

[28]        While the RSA was for up to a five-year term, it was clearly a revolving facility. Purchased receivables could be expected to be collected by MIH within about a month. MIH could be expected to know within a short period of time if any obligor’s payment history or prospects were declining or deteriorating, or if McKesson Canada’s creditworthiness was declining or deteriorating, and could take immediate steps to protects its interests and its future exposure, without waiting five years.

 

[29]        The RSA provided that McKesson Canada would pay MIH’s costs and expenses related to the transactions, including the costs of an inter-company transfer pricing study.[13]

 

[30]        The RSA provided that the purchase price payable by MIH to McKesson Canada for each receivable would be at a formulaically determined discount from its face amount. This is described in greater detail immediately below.

 

[31]        Any dilution or reduction to the face amount of a receivable resulting from discounts, rebates, disputes or returns, or by way of set-off under the terms of the receivable or otherwise, were deemed collections by McKesson Canada as servicer and to be accounted for to MIH as such. This did not apply to prompt payment discounts for reasons that were not directly explained in the evidence. The scope and nature of how the dilution risk relating to prompt payment discounts is dealt with under the agreements is in issue in this appeal.

 

[32]        During the term of the RSA it was discovered that interest received on the transferred receivables had not been accounted for and was retained by McKesson Canada. This was not consistent with the RSA.[14] The parties agreed not to account for the past error or to correct it going forward. In 2005, the parties agreed in writing that the interest and late payment charge obligations on the transferred receivables were never intended to be transferred – something that is difficult to reconcile with the RSA language or on any arm’s length basis.

 

[33]        The agreements are governed by Luxembourg law.

[34]        MIH, McKesson Canada’s direct parent, borrowed all of the money in Canadian dollars to purchase the receivables from an Irish company in the McKesson Group (alluded to above), that was one of its indirect parents. The purchase of receivables under the RSA was MIH’s stated use of the funds in its loan agreements. The interest payable by MIH was a direct function of the discount enjoyed by it under the RSA. MIH’s obligation to repay its borrowings to its Irish affiliate was fully guaranteed by its indirect parent, another related Luxembourg company (“MIH2”). In addition, MIH enjoyed an indemnity from MIH2 under a Memorandum of Understanding for any amounts payable in accordance with the RSA that were not received from McKesson Canada in order to allow MIH to fully re-pay its borrowings from its Irish affiliate.[15] MIH, McKesson Canada’s counter-party to the RSA and the Servicing Agreement, did not take any financial risk under this group of contemporaneous, inter-woven agreements all of which were financially and legally linked and related. All such risk was borne by other entities in the McKesson Group. That risk ultimately remained economically with McKesson U.S., everyone’s ultimate parent company, both before and after the RSA transactions.

 

a) The RSA Discount Formula

 

[35]        The amount payable for a purchased receivable under the RSA was the product obtained from multiplying (i) the face amount of the receivable and (ii) one minus the Discount Rate expressed to four relevant digits.[16] By way of illustrating the method in which this formula worked, if the Discount Rate (as defined itself by a further formula) worked out to 0.0150, MIH would pay $98.50 for every $100 of receivables[17], buying them at a 1.5% discount from face.

 

[36]        The Discount Rate is defined in the RSA to be the sum of (i) the Yield Rate, (ii) the Loss Discount, and (iii) the Discount Spread.

 

(i)      The Yield Rate

 

[37]        The Yield Rate was the 30 day Canadian dollar bankers’ acceptance (BA) rate, or CDOR,[18] on the first business day of the relevant settlement period. This operated as the floating base rate. This was intended to reflect a current, risk-free market rate of return. It is not challenged as an appropriate floating base rate for this Canadian dollar, Canadian obligor transaction.

 

                   (ii)     The Loss Discount

 

[38]        The Loss Discount was intended to reflect the credit risk of the McKesson Canada customers who were the receivables debtors. The Loss Discount was made up of two parts: (i) a Loss Discount component applicable to the Designated Obligors (whose receivables each exceeded 2% of the pool) and (ii) a Loss Discount component applicable to the other smaller Obligors making up the more diversified majority of the receivables pool.

 

[39]        The RSA expressed a fixed Loss Discount of 0.23% applicable to the initial purchase of receivables in 2002 to the end of 2003. This applied to the year under appeal.

 

[40]        For the remaining term, the Loss Discount was to be recalculated each year starting January 1, 2004. Also, if at any time MIH felt that the ratio of Designated Obligors’ receivables to other Obligors’ receivables in the pool was materially different than originally calculated for a year, MIH could require the Loss Discount to be recalculated that month. McKesson Canada as seller did not have any similar right. The result was that the RSA required the Loss Discount to be recalculated annually, and permitted only MIH to require it to be recalculated as often as monthly in the event of such a material change in risk.

 

[41]        For the larger Designated Obligors, a schedule to the RSA fixed their Individual Loss Discounts for the entire five-year period. In computing the aggregate Loss Discount for Designated Obligors, the weighted average of these fixed Individual Loss Discount amounts (weighted by each Designated Obligor’s share of the receivables pool as at the end of the prior year) was used. Since the Individual Loss Discounts were intended to reflect the unique credit risk of each Designated Obligor, there was a considerable range within the scheduled amounts (ranging from approximately 0.04% to 0.35% - a nine-fold range).

 

[42]        For each of the other Obligors (comprising about two-thirds of the pool), the RSA fixed an Individual Loss Discount at 0.2380% for the entire five-year term. The mechanics of how this number was arrived at was not apparent from the RSA as the agreement simply stipulated a fixed number for the entire duration. In computing the Loss Discount for these other Obligors, the weighted average of this fixed 0.2380% (weighted by the total of these other Obligors’ share of the receivables pool as at the end of the prior year) was used.

 

[43]        The Loss Discount was the sum of the weighted Individual Loss Discount for all Obligors.

 

[44]        The initial fixed Loss Discount of 0.23% applicable to the first full year was in fact calculated on this same basis. The amount thereof is significantly in issue in this appeal.

 (iii)   The Discount Spread

 

[45]        The RSA fixed the Discount Spread at 1.7305% for the entire duration of the agreement. Since this was fixed, the agreement does not describe how this number was arrived at. The evidence is that generally this relates to (i) the risk that McKesson Canada’s creditworthiness deteriorated significantly, and receivables debtors might set off their rebate entitlements in such event, (ii) the risk that McKesson Canada’s customers might increase their take-up of available prompt payment discounts, (iii) the risk that MIH might decide to appoint a new servicer following any termination event who might require a greater servicing fee than provided for successor servicers in the Servicing Agreement and (iv) the need for the Discount Rate to fully cover MIH’s cost of funds.

 

[46]        There was no corresponding consideration given to the possibility of McKesson Canada’s creditworthiness improving, customers taking less advantage of prompt payment discounts, or the impact of more prompt payments on the DSO. These imbalances were never explained.

 

[47]        This 1.7305% amount, it being a fixed amount, and the extent of MIH’s exposure to McKesson Canada credit risk under the agreements in the circumstances, are also significantly in issue in this appeal.

 

3.       The TDSI Opinions on Arm’s Length Terms & Conditions and Pricing

 

[48]        The RSA and all related agreements were first signed as of December 16, 2002. The conception, structuring, planning and drafting was under way for an unknown amount of time before that. This process seems to have been lead primarily by the Vice-President of Taxes of McKesson U.S., together with the tax and banking lawyers at Blake, Cassels & Graydon LLP (“Blakes”). Some general transfer pricing advice on approaches to, and issues in, structuring such a transaction was obtained in the summer of 2002 from Horst Frisch, a U.S. consulting company specializing in transfer pricing.[19] McKesson Canada’s role was limited to providing support and information regarding such things as its customers, its receivable portfolio, its projections, and its credit and collection policies et cetera.

 

[49]        In the weeks before the signing of the agreements, probably around December 1, 2002, Toronto Dominion Securities Inc. (“TDSI”) was retained by Blakes to provide advice on certain arm’s length aspects of certain of the terms and conditions of the RSA and of certain components of the discount calculation. The TDSI engagement letter was sent to TDSI on December 3^rd. It is clear from TDSI’s advice that by the time that TDSI was consulted, the structure and pricing approach and formulae were largely settled. It is not clear that any significant changes were made to these to reflect any advice or information given by TDSI. This is consistent with the testimony of Mr. Hooper of TDSI.

 

[50]        It can be noted that in 2002 the Act included contemporaneous arm’s length transfer pricing documentation/analysis requirements to defend the 10% transfer pricing penalty provisions. In fact, the TDSI opinions were relied on as the only contemporaneous basis to successfully contest CRA’s pre-reassessment proposal to impose transfer pricing penalties.[20]

[51]        TDSI’s advice was initially sought on (i) receivables eligibility criteria (ii) termination events/triggers and (iii) the appropriateness of the discount pricing. Somewhat oddly and not explained, Blakes’ engagement letter specifically identifies and raises the possible need to address the effect of a potential replacement servicer under the Servicing Agreement as part of the discount.

 

[52]        McKesson Canada had a pre-existing business relationship with the Toronto Dominion Bank group. The full scope of that was not put clearly in evidence, however, several years earlier, McKesson Canada had done a receivables factoring transaction with TD Factors. The TD Factors transaction was an entirely tax-driven year-end short-term transaction designed to avoid Canadian federal capital tax and seems to have been priced accordingly. It is entirely unreasonable to suggest this was a truly comparable transaction for arm’s length pricing purposes to the one in issue in this appeal.

 

[53]        Barbara Hooper is the person at TDSI that Blakes chose to contact. She was known to be a senior member of TDSI’s securitization group. Her advice was sought notwithstanding that everyone knew the RSA and related transactions were not structured as securitization transactions, were not intended to be securitization transactions, and that the purpose, objective and characteristics of the RSA transactions were significantly and materially different than a securitization transaction. No advice or information was sought from anyone other than the TDSI securitization group (and the TD bond traders briefly and casually consulted by Ms. Hooper’s group.)

 

[54]        Barbara Hooper was and is clearly a recognized professional and experienced expert (as a business person would use that word) in securitization matters, including trade receivables securitizations. She testified in this hearing as a material witness and not a qualified expert witness. Since her role as a material witness in offering the TDSI opinions involved her exercise of her professional judgment, she was allowed to fully explain in her testimony those opinions, including her reasons, and her supporting information, bases and subsidiary opinions.

 

[55]        Clearly Ms. Hooper’s experience with trade receivables securitizations qualified her to give valuable advice to the McKesson Group entities participating in the transaction and to the Court. She is certainly very knowledgeable about the risks associated with transfers of trade receivables, how those risks can be identified, and how those risks can be minimized in a receivables securitization transaction. Her experience and expertise did not however extend to pricing those risks if the risks were to be transferred, nor did it extend to market discount rates applicable to outright non-recourse or limited recourse sales or factoring of trade receivables. She did not hold out or suggest otherwise in the TDSI opinions or in her testimony.

 

[56]        Her testimony and TDSI’s involvement in the RSA transactions have been helpful to the Court. The TDSI opinion followed the conceptual approach dictated by the RSA presented to it by the McKesson Group. As is often the case with expert and other opinion evidence, the Court found much of the detailed explanation and reasoning behind the opinions, as well as some of the data and information supporting the opinions, helpful notwithstanding that the Court does not arrive at entirely the same conclusion in the end.

 

a) Eligibility Criteria

 

[57]        The TDSI opinion deals with this in a single short three-sentence paragraph. It concludes, without explanation or analysis, that the definition of eligible receivables in the RSA is within the range of normal in an arm’s length transaction of this nature. While it uses the words “transaction of this nature”, I can only conclude this is a reference to a securitization transaction involving receivables given the description of the experience TDSI brought to bear.

 

[58]        It goes on to identify that the exposure to the receivables pool concentration levels associated with McKesson Canada’s Designated Obligors would not be present in a securitization, will need to be addressed by TDSI in addressing the Discount Rate, and that this component of the Discount Rate will need to be dynamic, reflecting possible changes in the relevant balance of Designated Obligor receivables in the pool from time to time.

 

b) Termination Events

 

[59]        TDSI is satisfied that the triggers in the RSA definition of termination event are within the range of normal in an arm’s length transaction “of this nature”. I repeat my earlier observations about her use of this phrase in the TDSI opinion.

 

[60]        The TDSI opinion makes specific reference to the role of such termination triggers as protection for poor performance of receivables or declining creditworthiness of the seller. It identifies McKesson Canada’s creditworthiness as seller as relevant in part because of its obligations to remit collections to MIH. TDSI is of the express view that “because [McKesson Canada] is so closely tied and important to [McKesson U.S.], it is reasonable to use the public debt ratings of [McKesson U.S.] as an indication of [McKesson Canada’s] creditworthiness”.

 

[61]        The TDSI opinion goes on to specifically consider i) the receivables pool’s delinquency ratio trigger in the RSA, and ii) the receivables pool’s loss ratio trigger in the RSA.

 (i)     Delinquency Ratio Trigger

 

[62]        TDSI considered the improving two year historical trend in the delinquency ratio of McKesson Canada’s receivables and the recently maintained 1.0% rate. The 2.5% trigger rate in the RSA would, in TDSI’s opinion, represent a significant adverse deviation from the current steady state of 1% and so considered reasonable. TDSI highlighted the importance of the dynamic four-month rolling average approach to measuring the delinquency ratio in the RSA, and uses this approach in its analysis. TDSI confirmed that this is consistent with the three to six month periods generally used for such purposes.

 

(ii)     Loss Ratio Trigger

 

[63]        TDSI looked at three years of historic bad debt experience on McKesson Canada’s receivables portfolio. TDSI identified the difference between accounting write-offs and the 90 day delinquency definition of losses for purposes of the loss ratio in the RSA, with the result that the latter ratio could be expected to exceed the former. TDSI opined that a dynamic loss ratio, which measured a four-month average 90 day delinquency, and with a trigger of 0.25%, appeared reasonable given that, although write-offs to sales on a monthly basis at times reached this level, it had never exceeded 0.10% on a four-month rolling average.[21]

 

          c) Discount Rate

 

[64]        The TDSI Report says its assessment of the appropriate compensation was set by (i) where possible, looking at pricing of comparable risks in the market and (ii) “where pricing comparables were unavailable, we assessed the potential cost to [MIH] of assuming the risk”.[22]

 

[65]        In its summary paragraphs on the total discount at the end of its report, TDSI addressed the total discount following the RSA definition of Discount Rate as the sum of (i) the Yield Rate, (ii) the Loss Discount, and (iii) the Discount Spread. In summarizing the TDSI Report on the Discount Rate, it is easiest for the Court to follow this same order in its Reasons rather than the different order more loosely followed in the body of the TDSI Report.

 

(i)      The Yield Rate

 

[66]        TDSI identifies that the RSA used the 30 day Canadian dollar CDOR/BA Rate as the floating base rate component and that the 30 day CDOR Discount Rate needs to be adjusted to reflect the receivables’ DSO.

 

[67]        Notwithstanding the definitions of Discount Rate and Yield Rate in the RSA (even after it was amended and restated, and after being further clarified), the Court observes that the Yield Rate component of the Discount Rate in fact needs to be adjusted to reflect that the 30 day CDOR Rate is expressed as an annual rate and therefore needs to be adjusted to reflect the receivables pool’s DSO by multiplying it by the DSO and then dividing by 365.

 

[68]        TDSI selected the receivables pool’s three year average monthly DSO it calculated as 32. There was no discussion of why a shorter period’s average monthly DSO was not used, nor was there any discussion in the TDSI Report of a dynamic rolling average approach to DSO.

 

[69]        TDSI recognized that the DSO for the initial purchase of approximately $460,000,000 of receivables would very significantly overstate the expected payment term for these receivables as they were a pool of mature short-term receivables (i.e. some would be paid the following day because they had been outstanding for between 1 and 30 days or more).

 

[70]        TDSI’s approach was to instead use a 16 day DSO for the initial purchase. However, it did not in fact calculate the Discount Rate for the initial December 16, 2002 purchase using a 16 day DSO; it instead averaged the “missing” 16 days across the entire five years of the RSA term, and set a fixed DSO of 31.73 days for all Discount Rate calculations for all receivables purchases under the RSA. It is obvious that the effect of this was to double this portion of the Discount Rate for the December 2002 purchase thereby providing a significant timing benefit to the McKesson Group in respect of McKesson Canada’s tax reduction for the 2003 year in issue.

 

(ii)     The Loss Discount

 

1.       Designated Obligors

 

[71]        TDSI begins its Loss Discount analysis by determining the monthly write-offs to sales from the 3 to 4 years of data provided to it for McKesson Canada, together with the same numbers computed on a three and 12 month rolling average basis. They conclude that, viewed from this dynamic perspective, McKesson Canada’s write-offs to sales are “very low”.

 

[72]        TDSI identifies a concentration risk issue associated with the larger Designated Obligors that losses on their receivables have an increased likelihood of deviating from historical levels (presumably not necessarily adversely). TDSI does not attempt to quantify that increased likelihood, nor most importantly, does it analyze the historic data associated with Designated Obligors separately to try to validate the increased risk.[23]

 

[73]        TDSI looked at the Designated Obligors separately from the other Obligors and looked at each Designated Obligor individually (treating all hospitals as a single Designated Obligor), as contemplated by the RSA.

 

[74]        TDSI thought it appropriate to consider each Designated Obligor’s public debt rating, or if a Designated Obligor did not have one of its own, its parents’ rating without adjustment. If neither was rated, the TDSI Report assumed a non-investment grade credit rating. TDSI then looked at each Designated Obligor’s 180 day risk credit spreads in the public debt market. TDSI thought it was appropriate to treat MIH’s risk under the RSA as 180 day risk which would allow for TDSI’s estimate that it would take 90 days to wind down the pool and liquidate the portfolio upon termination.

 

[75]        TDSI estimated the Individual Loss Discount for each Designated Obligor using this market based discount, first adjusted for the DSO and then applied to (multiplied by) its total receivables at that time (December 2002). These numbers were then fixed in the RSA’s Schedule D of Designated Obligors’ Individual Loss Discounts.[24]

2.       Other Obligors

 

[76]        The TDSI Report uses one month historic data to assess the Loss Discount attributable to the smaller Obligors comprising the majority of the receivables pool. The report gives no reason for choosing the one month historic average instead of either the three or 12 month rolling averages also computed and set out in this portion of the TDSI Report (or, for that matter, the four month rolling average set out and used by TDSI in the portion of its report opining on loss trigger termination events). The TDSI Report does not appear to even break out the Designated Obligor portion of the pool for this purpose to look only at the other Obligor performance. Neither the report nor Ms. Hooper’s evidence provided any explanation.

 

[77]        TDSI goes on to note that the multi-year data it has been provided does not cover a full economic cycle of Canadian trough-to-peak-to-trough. The report does not say more historic data was asked for, nor that it was unavailable. For this reason, it “suggests” adding three standard deviations from mean (although TDSI only set out average not mean in its charts). No reason is given for this suggestion that three standard deviations be used. This suggestion, combined with the selection of monthly numbers, had the result of driving the .03% actual monthly average loss to 0.24% - that is an eight-fold increase. (Though the TDSI Report then uses 0.2380%; presumably more accurate four digit calculations had been done outside the report).

 

[78]        It can be noted from TDSI’s chart that, while the computed standard deviation was .07% for monthly historic averages, it was a fraction of that for the rolling 12 month and three month historic loss performance (.02% for 12 month and .04% for three months). It can be observed readily that had this same approach of adding three standard deviations to twelve month historic rolling average losses been taken, the twelve month average of .04% would only increase to .1% - which would still be a two and one-half fold increase (and still without any given reason for using three standard deviations).[25]

 

[79]        TDSI correctly points out that this approach to the Loss Discount percentage applicable to other Obligors is a function of losses on the dollar amount of receivables, not an annual or other rate that is a function of time, and thus does not need to be adjusted for the DSO.

 

[80]        The TDSI Report then simply totals the pool’s weighted DSO adjusted credit spread for each Designated Obligor consistent with the RSA, and the other Obligors’ 0.2380% weighted to their share of the receivables pool owing by the other Obligors.

 

[81]        TDSI described the annual (or earlier at MIH’s request) adjustment of the Loss Discount component of the Discount Rate as appropriate. It did not discuss why it considered those appropriate.

 

(iii)    The Discount Spread

 

[82]        The set and fixed discount spread in the RSA of 1.7305% appears from the TDSI Report to have been built up by TDSI. The TDSI Report built up this number from four components (1) a servicing discount, (2) a prompt payment dilutions discount, (3) an accrued rebate dilutions discount, and (4) a credit spread interest discount.

 

[83]        It is not at all clear from the RSA, the TDSI Report, or the witnesses why these four risks are addressed by something called a discount spread. In any event, it is clear that the parties to the RSA and related transactions, and their advisors, considered these the material risks associated with the RSA transactions other than the risk of loss on the receivables themselves resulting from Obligors’ financial defaults.

1.       Servicing Discount

 

[84]        The TDSI opinion begins its analysis with the points that (i) the RSA requires MIH to service the purchased receivables, (ii) under the Servicing Agreement the servicer (including the initially appointed servicer, McKesson Canada) will be paid a fixed fee of $9,600,000 annually, paid monthly, to service the receivables pool regardless of size, and (iii) the Servicing Agreement contemplates the prospect of MIH choosing to, or needing to, appoint a replacement servicer.

 

[85]        TDSI ties the “very possible” need for a replacement servicer to the fact that MCC is not a highly-rated credit on a stand alone basis. The connection is not described further and is not obvious.

 

[86]        TDSI then set out the replacement servicer pricing it obtained from a single source (believed to be perhaps a major accounting firm). This indicated collection fees for current accounts in the range of 1.0% to 3.0% of the face amount of the receivables. TDSI did not seek replacement servicer pricing from TD Factors.

 

[87]        TDSI observed again that a replacement servicer would only be needed for a short period of time once the agreement was terminated and no further receivables were purchased by MIH.

 

[88]        TDSI then chose to use a 2% replacement servicer fee for the reason that it was the midpoint of the 1% to 3% range.

 

[89]        TDSI observed that the approximately $800,000 fee payable monthly under the Servicing Agreement would be in the range of 0.2% assuming that approximately one-half of the $900,000,000 cap, or about the original purchased amount of receivables, was outstanding throughout the term. No attempt is made in the report to explain or substantiate the “expected sales” volumes.[26] No explanation is made of the relevance of a $900,000,000 cap that does not relate to expected sales. No attempt is made to explain the tenfold difference between the 0.2% charged by McKesson Canada and TDSI’s selection of 2% from the range of 1% to 3% for a replacement servicer. Nor is this last issue dealt with in TDSI’s supplemental report.

 

[90]        The TDSI Report then notes that a Moody’s publication consulted by it indicated that, in the previous year, 9.41% of companies with McKesson U.S.’ rating were downgraded to the RSA trigger rating or had their rating withdrawn. From there, without any further explanation, TDSI then used a “conservative” “assumption” of a 25% “chance” that a replacement servicer would need to be appointed.[27]

 

[91]        Using these numbers alone, TDSI then multiplies 2.0% times 0.25, and 0.2% times 0.75, and adds these to arrive at a “reasonable expected cost of servicing” of 0.65%. This number, three and one-quarter times what McKesson Canada or a replacement servicer is to be paid to service the receivables under the Servicing Agreement, is then used to discount each receivable (including the $465,000,000 worth purchased upon signing the agreement which appeared to have virtually every likelihood of being paid in orderly fashion beginning the very next day, and including those purchased on the next day, and the day after that, et cetera).

 

[92]        Several months after the transactions were in place, TDSI was asked for follow-up advice on the fixed fee payable to the servicer under the Servicing Agreement. TDSI’s supplemental report is described further below.

 

2.       Prompt Payment Dilutions Discount

 

[93]        As described above, the RSA does not treat prompt payment discounts enjoyed by McKesson Canada’s customers as deemed collections on their receivables. The TDSI Report accepts this without question and without even addressing whether this would be normal in arm’s length securitization transactions. It is not self-evident why prompt payment discounts are not treated in the same manner as other dilutions such as volume rebates to customers.

 

[94]        Thus, TDSI identifies the risk in the RSA that there may be a change in the extent to which McKesson Canada’s customers take up the prompt payment discounts offered by paying their receivables earlier than past individual or overall experience might suggest. For example, if MIH purchases $100 of receivables and the purchaser promptly pays $98 in full satisfaction, McKesson Canada does not have to account to MIH for the $2. The RSA pricing therefore had to somehow address the fact that prompt payment discount rights would be exercised by McKesson Canada customers. MIH would be overpaying if the embedded take-up was underestimated and McKesson Canada would be underpaid if the embedded take-up was overstated. TDSI computed McKesson Canada’s prompt payment discounts historically on an annual basis as 0.5% of sales and observed this is “very consistent”.[28]

 

[95]        TDSI then suggested that a 20% “buffer” be added to bring the 0.5% to 0.6%. The TDSI Report does not attempt to explain why a 20% buffer was chosen. It does not acknowledge, much less address, the fact that if up to 20% more customers start enjoying prompt discount payments on the terms already offered, there would be a corresponding favourable impact on the DSO of the receivables pool which would have a correspondingly material impact on risk and pricing as MIH would be paid materially more quickly.

 

[96]        Unstated and unexplained, but implicit in the TDSI Report, is that TDSI is fine with the prompt payment discount risk being estimated at the outset, fixed throughout at historic levels, plus an unexplained buffer, instead of it being reflected in its rolling average actual performance throughout the RSA’s term, and even though it is to be incorporated in the MIH RSA risks instead of being treated as a deemed collection.

 

3.       Accrued Rebate Dilutions Discount;

 

[97]        McKesson Canada’s volume rebates are paid separately to its customers and are not enjoyed at point of purchase or time of payment. Rebates are instead paid directly by McKesson Canada to its customers periodically. The risk that transferred receivables would be subject to a reduction for rebates was therefore not transferred to MIH nor accounted for under the RSA. However, there was nonetheless the risk that a customer would choose to set-off its McKesson Canada payables by the amount of an anticipated or earned but yet unpaid rebate owing to it from McKesson Canada. TDSI speculated this may happen if customers became concerned about McKesson Canada’s financial situation.[29] The RSA expressly and clearly requires such a set-off risk to be fully borne by, and indemnified by, McKesson Canada. Thus, TDSI identifies that the RSA has MIH taking McKesson Canada (MIH’s direct subsidiary) credit risk for McKesson Canada’s indemnity obligation.

 

[98]        TDSI computed the three-year historic rebates as a function of sales numbers. The average was 3.8%. The minimum was 2.13% and the maximum was 5.53%. Each of these percentages is a total of all accrued rebates. Thus, without further adjustment, using these raw numbers would reflect the situation where all of McKesson Canada’s customers decide it is time to set-off rebates. Without any explanation other than to be conservative, TDSI selected the highest maximum number of 5.53% of the receivables balance as the amount at risk for set-off. No suggestion is made by TDSI that a dynamic rolling average approach should be taken or considered through the term of the RSA. TDSI does not address the prospect that MIH might be expected to have and exercise termination rights if McKesson Canada’s financial situation were to become such that MIH was concerned about its financial viability. The TDSI Report does not say they are aware that any such set-off has ever been made, claimed, threatened or even asked for.

 

[99]        To this maximum set-off amount, TDSI applies a factor selected by it to reflect McKesson Canada credit risk. For this purpose they assessed a 5.25% credit spread (that is the amount charged above a lender’s floating fixed base rate) for McKesson Canada on the basis that, without its own credit rating, they should use non-investment grade credit spreads as a proxy. (Non-investment grade borrowers’ bonds in the bond market are also referred to as high yield bonds or junk bonds). This, even though several times elsewhere in the TDSI Report, they use unrated Obligors’ parent ratings as their proxy, and accept McKesson U.S.’s credit rating risk as reflective of McKesson Canada’s creditworthiness risk in the discussions of termination event and triggering event. The TDSI Report does not explain the relationship of its estimated 5.25% credit spread to McKesson Canada’s actual rates available to it from its available lenders and facilities which, as described elsewhere, are very different. Notably, given that the RSA clearly did not require McKesson Canada to segregate collections (absent a termination event), the McKesson Group and MIH had little apparent concern about McKesson Canada’s creditworthiness or financial situation. After all, the risk of collections not being passed through to a purchaser of receivables is one of the significant risks in a factoring transaction.

 

[100]    The TDSI Report does not address MIH’s rights under the RSA to have the funds segregated and not commingled at any time there has been a termination event including a material adverse change. Nor does it address whether the failure to segregate absent a termination event and the willingness to accept commingling risk in the RSA about termination is consistent with market practices in arm’s length negotiated agreements such as those involving securitizations or arm’s length servicers.

 

[101]   TDSI multiplies the 5.53% maximum historic set-off risk by the 5.25% non-investment grade borrower credit spread (adjusted for the DSO because credit spreads are expressed as an annual rate) to arrive at a fixed discount of 0.0244% for the accrued rebate dilutions discount for the entire term of the RSA. There is no explanation offered for accepting a fixed approach.

 

4.       Interest Discount

 

[102]   This aspect is dealt with in a single paragraph of the report. It is not a risk- based assessment but reflects TDSI’s observations that “in addition to being compensated for the risks it is assuming, [MIH] must also cover its cost of capital out of the discount.”

 

[103]   TDSI then says that because MIH is “exceptionally thinly capitalized”[30] it is again appropriate to use non-investment grade bond market credit spreads as a proxy.

 

[104]   These statements may be correct from a profitable business transaction point of view, however TDSI does not address the possibility that the transaction exactly as structured could not be done profitably on arm’s lengths terms and conditions, or the fact that cost of funds/source of financing is an expense to a buyer that does not generally increase or decrease the value of a seller’s assets, or any other issues raised thereby.

 

[105]   TDSI was not made aware of MIH’s sources of financing, and was therefore unaware it also enjoyed a full guarantee and indemnity from its parent company in the McKesson Group.

 

[106]   The TDSI Report then goes on to add a five-year swap spread for non-investment grade bond (i.e. junk)  issuers of 5.25% (as described above with respect to rebates) to the 30 day CDOR rate because of the five-year term of the agreement.

 

[107]   This non-investment grade/high yield/junk bond rate is thus factored into the discount pricing twice by TDSI, once to reflect McKesson Canada’s creditworthiness with respect to the maximum rebate exposure potential, and again to reflect MIH’s needed return to cover its cost of the full amount borrowed by it.[31]

 

[108]   TDSI does not explain why it uses five year pricing for funds needed by MIH in a monthly settled revolving facility with a DSO of about a month and which TDSI estimates (twice) in its report as perhaps as much as 180 day exposure (which includes a 90 day allowance to wind up).

 

[109]   As with the other components of the Discount Spread, the TDSI Report does not address the appropriateness of the Interest Discount credit spread being at a fixed rate throughout the five-year term.[32]

 

[110]   Overall, it appears clear that each of the four components of the Discount Spread have clearly been computed using the maximum numbers even arguably justifiable (not necessarily being, or clearly being, the maximums within a reasonable range), and then fixed each of those at those maxed out numbers for the entire five-year term without adjustment or recalculation on any basis even though there are dynamic monthly and annual components (some of which are very similar) and material adverse change or MAC rights elsewhere in the RSA agreements. This lack of balance in favour of MIH is not expressly identified, and not discussed at all in the TDSI reports or evidence.

 

4.       TDSI’s Supplemental Report on Servicing Fees

 

[111]   In April 2003, after the execution and implementation of the RSA and related transactions, TDSI was asked to report to McKesson Canada, MIH and Blakes on whether the fixed monthly fee of $800,000 paid under the Servicing Agreement to administer and collect all the receivables in the pool was within the range of normal for this type of arrangement.

 

[112]   TDSI again begins by being clear that its expertise in this area is based upon TDSI’s experience arranging Canadian trade receivables securitization transactions, which are structured without a separate servicing fee being paid or negotiated and “accordingly, we do not have ready knowledge of comparable situations”.

 

[113]   The TDSI servicing report carries on:

 

The level of the servicing fee must be assessed relative to the resources required to effectively service the portfolio. The resource requirements will depend upon the size and composition of the receivables portfolio.

 

[114]   It then notes, however, that under the Servicing Agreement, McKesson Canada is paid a fixed fee each month regardless of the size of the portfolio. TDSI computes that the fixed annual fee at $9,600,000 works out to somewhat in excess of a 1% per annum fee on the $900,000,000 limit and somewhat in excess of a 2% fee on the then current $460,000,000 range pool. TDSI’s investigations turned up servicing fees of 1% per annum recorded by Bell Canada in a $1,000,000,000 receivables securitization and 2% per annum by Telus in a $650,000,000 receivables securitization. TDSI notes that these receivables would involve primarily Telus and Bell’s retail customers’ phone bills.[33]

 

[115]   TDSI then looked at the Servicing Agreement fee as a function of the dollars to be collected and determined this was approximately 0.1% of the maximum eligible pool size of $900,000,000 and a 32 day DSO, and approximately 0.2% based on the receivables balance at closing. TDSI’s investigations found only one U.S. transaction with relevant information disclosed publicly which reflected a servicing fee equal to 0.1% of collections.

 

[116]   Finally, TDSI contacted two large rating agencies with considerable involvement in rating securitization transactions. According to Moody’s, negotiated servicing fees between buyers and sellers are typically in the range of 1% to 2% of the average receivables balance. According to Standard & Poor’s, servicing fees are usually 1% per annum of the average receivables balance. Note that both rating agencies describe it as a dynamic fee, based on average receivables balances during the year.

 

[117]   In its closing opinion, TDSI again notes it is unusual that the servicing fee is fixed. Based on the amount of receivables transferred at closing, TDSI concludes that the fee is somewhat higher than the few comparables. It is clear that by somewhat, TDSI meant up to 100%.

 

[118]   TDSI’s servicing opinion is dated April 25, 2003. The 2003 year under appeal ended in March 2003. I assume from the TDSI Report and discussion that the size of the receivables pool did not increase materially between the mid-December 2002 initial closing and the end of the year under appeal in March 2003.

 

5.       The Law

 

Subsection 247(2) of the Act provides as follows.

 

(2)   Transfer pricing adjustment -- 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	2) Redressement -- Lorsqu'un contribuable ou une société de personnes et une personne non-résidente avec laquelle le contribuable ou la société de personnes, ou un associé de cette dernière, a une lien de dépendance, ou une société de personnes dont la personne non-résidente est une associé, prennent part à une opération ou à une série d’opérations et que, selon le cas :
(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	a) les modalités conclues ou imposées, relativement à l’opération ou à la série :
(b) the transaction or series	b) les faits suivants se vérifient relativement à l’opération ou à la série :
(i) would not have been entered into between persons dealing at arm's length, and	(i) elle n’aurait pas été conclue entre personnes sans lien de dépendance,
(ii) can reasonably be considered not to have been entered into primarily for bona fide purposes other than to obtain a tax benefit,	(ii) il est raisonnable de considérer qu’elle n’a pas été principalement conclue pour des objets véritables, si ce n’est l’obtention d’un avantage fiscal,
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,	Les montants qui, si ce n’était le présent article et l’article 245, seraient déterminés pour l’application de la présente loi quant au contribuable ou la société de personnes pour une année d’imposition ou un exercice font l’objet d’un redressement de façon qu’ils correspondent à la valeur ou à la nature des montants qui auraient été déterminés si :
(c) where only paragraph (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	c) dans le cas où seul l’alinéa a) s’applique, les modalités conclues ou imposées, relativement à l’opération ou à la série, entre les participants avaient été celles qui auraient été conclues entre personnes sans lien de dépendance ;
(d) where paragraph (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.	d) dans le cas où l’alinéa b) s’applique, l’opération ou la série conclue entre les participants avait été celle qui aurait été conclue entre personnes sans lien de dépendance, selon des modalités qui auraient été conclues entre de telle personnes.

 

[119]   The Minister’s reassessments rely upon paragraphs 247(2)(a) and (c) to make a transfer pricing adjustment in respect of the RSA. These subparagraphs apply if a taxpayer (McKesson Canada) and a non-resident person with whom the taxpayer does not deal at arm’s length (MIH) are participants in a transaction[34] (each of the RSA and the Servicing Agreement) or a series of transactions (the RSA, the Servicing Agreement, and MIH’s loan agreement with its Irish indirect parent, the related guarantee thereof from MIH2 to the Irish company, and the related indemnity of MIH2 in favour of MIH regarding McKesson Canada’s obligations under the RSA), and the terms or conditions thereof differ from those that would have been made by arm’s length persons. If the “terms and conditions” do so differ, then the “amounts” that would otherwise be used by the taxpayer for purposes of the Act shall be “adjusted” to the “quantum or nature” of the amounts that would have been determined had the “terms and conditions” been those that arm’s length parties would have agreed to.

 

 (a) GlaxoSmithKline

 

[120]   The Supreme Court of Canada in GlaxoSmithKline had occasion to address the scope of the review of the relationships and circumstances that a Court is to undertake in a transfer pricing appeal:[35]

 

(1) A judge is to take into account all transactions, characteristics and circumstances that are relevant (including economically relevant) in determining whether the terms and conditions of the transactions or series in question differ from the terms and conditions to which arm’s length parties would have agreed.

 

(2)     The transfer pricing provisions of the Act govern and are determinative, not any particular methodology or commentary from the OECD Guidelines, or any source other than the Act.

 

I would add the observation that OECD Commentaries and Guidelines are written not only by persons who are not legislators, but in fact are the tax collection authorities of the world. Their thoughts should be considered accordingly. For tax administrators, it may make sense to identify transactions to be detected for further audit by the use of economists and their models, formulae and algorithms. But none of that is ultimately determinative in an appeal to the Courts. The legal provisions of the Act govern and they do not mandate any such tests or approaches. The issue is to be determined through a fact finding and evaluation mission by the Court, as it is in any factually based issue on appeal, having regard to all of the evidence relating to the relevant facts and circumstances.

 

(3) Arm’s length prices are established having regards to the independent interests of each party to the transaction. In this appeal, this means that the RSA transactions must be looked at from both the perspectives of McKesson Canada and of MIH.

 

(4) Other arm’s length transactions can be relied upon as comparables in a transfer pricing analysis only if either there are no material differences that would affect pricing, or if reasonably accurate adjustments can be made to eliminate the effects of such differences.

(5) Quoting from GlaxoSmithKline:

 

61        As long as a transfer price is within what the court determines is a reasonable range, the requirements of the section should be satisfied. If it is not, the court might select the point within a range it considers reasonable in the circumstances based on an average, median, mode, or other appropriate statistical measure, having regards to the evidence that the court found to be relevant. I repeat for emphasis that it is highly unlikely that any comparisons will yield identical circumstances and the Tax Court judge will be required to exercise his best informed judgment in establishing a satisfactory arm’s length price.

 

(b) Reasonableness

 

[121]   While GlaxoSmithKline involved the former subsection 69(2) transfer pricing rule, which was worded differently, I see no compelling reason to depart from the Supreme Court of Canada’s approach and comments in GlaxoSmithKline. While section 247 does not use the words “reasonable in the circumstances” or “fair market value”, arm’s length persons should generally be assumed for purposes of section 247 to act neither irrationally nor unreasonably having regard to all relevant circumstances. Similarly, arm’s length persons should generally be expected to transact for products and services at amounts within the range of their fair market value having regards to all relevant circumstances. This is not inconsistent with the wording of section 247.

 

(c) Relevant Series of Transactions

 

[122]   The Supreme Court of Canada in Copthorne Holdings Ltd. v. The Queen, 2011 SCC 63, mandates an expansive approach to the issue of series, given the inclusive nature of the meaning to be given to “series of transactions” in subsection 248(10). The starting point is the common law series in which each transaction in the series is pre-ordained to produce a final result. Then, subsection 248(10) deems any related transaction completed in contemplation of a series to be part of that series.[36]

 

[123]   The determination of the existence of a series and its constituent transactions is a question of fact to be determined on a balance of probabilities. While only amounts under the RSA have been challenged, clearly the Servicing Agreement, MIH’s Loan Agreement, and MIH2’s guarantee and indemnity are also all part of a series of transactions that included the RSA. In addition, their existence and terms also each meet the threshold of being relevant to a consideration of the RSA as described by the Supreme Court of Canada in GlaxoSmithKline. Each of these agreements are contemporaneous and interdependent in fact and in law. These agreements were needed to fund and to repay the RSA between McKesson Canada and MIH. They relate directly to the trade receivables of McKesson Canada’s business. McKesson Canada may not be a party to each of these agreements, but they each expressly refer to the terms and conditions of the RSA, or the sale, servicing, or collections on those receivables, et cetera.

 

[124]   Given that all of these transactions meet the threshold test of relevance in GlaxoSmithKline for purposes of considering the transfer pricing within the RSA itself, and given that the Crown has only challenged the amount of the discount used in the RSA, the Court does not need to rely further on the fact that these transactions together constitute a series of transactions for purposes of section 247.

 

 (d) Scope of Adjustments Permitted Under Paragraph 247(2)(c)

 

[125]   A reassessment under subparagraphs 247(2)(a) and (c) does not permit a recharacterization of the transactions entered into by non-arm’s length parties, nor can another different transaction entirely be substituted therefor. This would only be permitted under subparagraphs 247(2)(b) and (d) which have not been pleaded and the Crown is not relying upon. A transfer pricing recharacterization is only permitted under those provisions if arm’s length parties would not have entered into the transaction chosen by the non-arm’s length parties even with different terms and conditions and amounts, and if the only bona fide primary purpose of the transaction was to obtain a tax benefit.

 

[126]   However, it is clear from the provisions of section 247 that under subparagraphs (a) and (c) the Court is not limited to making adjustments with respect to the quantum of an amount in a term or condition that incorporates an amount. I do not accept the taxpayer’s submission that I am so limited. Paragraph 247(2)(a) is triggered when terms or conditions differ from those terms and conditions that arm’s length parties would agree to. There is no such limiting restriction on the phrase terms and conditions. Paragraph 247(2)(c) then mandates an adjustment to the quantum or nature of an amount used by the taxpayer for purposes of the Act to reflect the quantum or nature of that amount that would have been used had the “terms and conditions” conformed to what arm’s length parties would have agreed to.

 

[127]   Perhaps there is a point at which the extent of changes to the agreed non-arm’s length terms and conditions needed to reflect arm’s length terms and conditions in a transaction can constitute an effective recharacterization of the transaction only permitted to be affected under paragraph 247(2)(d) and only in the circumstances described in paragraph 247(2)(b) which provisions are not engaged in this appeal. Perhaps there also may be some terms and conditions in a transaction that are so fundamental that any particular change thereto could constitute in effect a recharacterization of the transaction. The Court does not need to venture anywhere close to that line in disposing of this appeal. That can be left for another day. In this case the Court is able to limit itself to a consideration of terms and conditions which it finds to not be on arm’s length terms and that directly relate to pricing.

 

(e) Factors that Exist Only because of the Non-Arm’s Length Relationship

 

[128]   Within a transfer pricing review, the question arises whether factors that exist only because of the non-arm’s length relationship are assumed away in the notional arm’s length analysis or remain relevant characteristics and circumstances.

 

[129]   This question may not arise to any extent in the context of a single purchase at a fixed price. The question does appear significant in the context of a long-term commitment to do things over a period of time. For example, in transactions such as those involving the RSA, does the Court assume a notional arm’s length MIH would still enjoy the benefit of the Irish company loan supported by the MIH2 guarantee and indemnity? In looking at transactions like the RSA, does the Court assume the notional arm’s length MIH still has the power throughout the term of the notional arm’s length contract to change McKesson Canada’s name, sell McKesson Canada, or do something else in order to trigger a termination event at will? Does the Court assume that the notional arm’s length purchaser still has the right to cause McKesson Canada to agree to change terms as they apply to future transactions under the agreement? Does the Court assume that the notional arm’s length MIH still has access to all of the financial information of McKesson Canada and information regarding its receivables portfolio and its entire business even though it may not be specified or required in the RSA?

 

[130]   This issue was addressed by Justice Pizzitelli in Alberta Printed Circuits v. The Queen, 2011 TCC 232:

 

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 …

 

… In General Electric, the Federal Court of Appeal confirmed that no error of law was made in taking into consideration the Appellant in that case, as a sub of its larger parent company, stood in the position of having an implicit guarantee by its parent of its bank debts.[37]

 

[131]   Based on this, all circumstances, including those that arise from, derive from or are rooted in the non-arm’s length relationship should be taken into account.

 

[132]   I think the better view is therefore that the Court can and should consider notional continued control type rights in appropriate circumstances when looking at term or executory contract rights. Not to do so would be to not look at all of the relevant characteristics and circumstances of the relationships. If these were to be ignored by a Court, companies within wholly controlled corporate groups could enter into skeletal agreements conferring few rights and obligations to the non-resident participant, (such as financial information disclosure, use of funds, financial covenants et cetera), all with the view to obtaining a more favourable transfer price to reduce Canadian taxes. Not approaching this issue this way would seem entirely inconsistent with this Court and the Federal Court of Appeal in G.E. Capital having focused on implicit unwritten, unenforceable guarantees of the parent company of the borrower. However, in this case, I do not need to do so in order to fully dispose of the appeal with respect to the proper transfer pricing adjustment, as detailed below. This too can be left for another day.

                                            

(f) The Rule in Browne v. Dunn and Opinions Within a Witness’ Expertise

 

[133]   The rule in Browne v. Dunn, (1893), 6 R. 67 (H.L.) generally requires counsel to give notice to those witnesses that the cross-examiner intends later to impeach. This allows the witness the opportunity to explain, and reflects a sense of fair play and fair dealing with the witness and his or her evidence.

 

[134]   In response to an objection by Appellant’s counsel, the Court had to rule in the course of the hearing on the applicability of the rule in Browne v. Dunn to the opinion and supporting evidence of the Appellant’s experts in respect of which the testimony of one of the Respondent’s experts expressed different views. The Court again had to rule on the same issue in response to a further objection by Appellant’s counsel as it related to one of the Respondent’s experts having different views from the TDSI Report and Ms. Hooper’s testimony regarding her opinion and her reasoning expressed therein. In both cases the Court denied the Appellant’s objection to the evidence of the Respondent’s expert.

 

[135]   The rule in Browne v. Dunn is not a fixed absolute. The extent of its application is within the discretion of the trial judge after considering the circumstances of the particular case per the Supreme Court of Canada in R. v. Lyttle, [2004] 1 S.C.R. 193 at paragraph 65.

 

[136]   The rule in Browne v. Dunn may understandably have different practical application in the particular case of material witness on a question of fact, than in the particular case of an expert witness on his or her opinion evidence and information supporting that opinion, and on the contents of the expert’s rebuttal report responding to the other party’s expert reports. In this regard, I would add that Ms. Hooper’s testimony relating to the TDSI Report was effectively testimony regarding her professional opinions within her area of expertise, and I would note that the opinions, reasoning and supporting information set out in the TDSI Report were the subject of considerable comment in the principal and rebuttal reports of the Respondent’s experts.

 

[137]   At the outset of the hearing it was agreed that each expert would testify as to the contents of both their principal report and their rebuttal report when called, notwithstanding that the Appellant’s rebuttal reports responded to principal expert reports of the Respondent’s experts before the Respondent’s experts had a chance to explain them to the Court. There was a right to recall any witness in the event this approach was considered to be found wanting on any point.

 

[138]   There could have been no doubt in the Appellant’s mind, nor the minds of its experts, that the Respondent’s experts challenged the opinions and reasoning of the Appellant’s experts and the TDSI Report.

[139]   In these circumstances of this case, there was absolutely no element of attempted surprise, much less any attempt to deny anyone an opportunity to fully explain. This was fully covered off by an opportunity to recall any witness if a party felt further explanation was needed. This was a standing offer that the Appellant chose not to take advantage of, even after the offer was re-extended as part of denying the objections.

 

[140]   What was in issue in the circumstances of each of the objections raised in this particular case related to the witness’ professional opinions without any suggestion or inference that he or she was not speaking the truth or was a witness unworthy of credit. This was in reality only a matter of competing and differing professional opinions.

 

[141]   This approach to the resolution of these objections finds support in R. v. Union Carbide Canada Ltd., [1991] O.J. 1213 (Ont. Court of Justice).

 

[142]   An attempt to force blind adherence to the rule in Browne v. Dunn should not be allowed to interfere with an orderly ordering of witnesses. Nor should it require that further rebuttal reports be filed in response to rebuttal reports when the mischief addressed by the rule is not present in the particular circumstances.

 

(g) The Court’s Analytical Approach to be Followed in this Case

 

[143]   The Court’s analysis of the pricing of the Discount Rate in the RSA will proceed on the basis that transactions described above are the transactions McKesson Canada and others in the McKesson Group chose to enter into. The real task in this transfer pricing appeal is for the Court to determine whether or not the terms and conditions of the transactions resulted in a Discount Rate committed to in the RSA by McKesson Canada that was within the range of what McKesson Canada and MIH would have agreed to had their transactions used terms and conditions that affect the pricing of the Discount Rate which persons dealing at arm’s length would have used.

 

6.       The Position of the Appellant

 

[144]   The Appellant’s position with respect to the transfer pricing adjustment can be generally summarized by me as follows:

 

Methodology:

 

[145]   The appropriate transfer pricing analysis methodology is not one of the four methodologies named in the OECD Guidelines and, therefore, an “other method” should be used. The taxpayer called an expert witness, Dr. Horst Frisch of Horst Frisch, in support of this view.

 

[146]   Based upon the evidence presented, I agree. (I am not sure that the Respondent substantially disagreed).

An OECD “other method”:

 

[147]   An “other method” that would be appropriate is precisely that followed by Barbara Hooper of TDSI, set out in the TDSI Report, and explained in her testimony at trial.[38] This analysis takes the RSA as the parties structured it and does not introduce or rely on any recharacterization or additions to the RSA. She expertly and accurately identified the risks transferred to MIH with its purchase of the receivables under the RSA. The TDSI analysis and report was contemporaneous. While Ms. Hooper did not testify as an expert witness, she and TDSI clearly had considerable expertise in trade receivables transfers in a securitization setting. That allowed an accurate identification of the risks inherent in the terms and conditions of the RSA. She was able to use her experience and knowledge in reviewing and commenting upon the pricing of the Discount Rate in accordance with the terms and conditions of the RSA.

 

[148]   I agree that this is an appropriate “other method” to be considered in reviewing the terms and conditions of the RSA to determine whether those that impact on pricing under the RSA are on arm’s length terms.

 

Another OECD “other method”:

 

[149]   Another “other method” that would be appropriate was that set out in the expert report of Mr. Jeremy Reifsnyder. This approach did not follow the parties’ chosen structure set out in the RSA. This approach relied on a comparability and adjustment approach which began by looking at the public bond market discount rates on a five-year non-investment grade Canadian bond fund index and then making a number of significant adjustments to that to reflect certain issues and characteristics particular to the RSA. It is fair to say that the use of this “other method” was the Appellant’s principal position.

 

[150]   The Reifsnyder methodology is described in greater detail below along with my reasons for finding it would not be appropriate to follow it as an acceptable “other method” in this case. That is not to say that parts of Mr. Reifsnyder’s analysis and the opinions that he expressed, and information that he relied on, were not helpful to the Court. Indeed, as described below, apart from (i) his build up being based upon an index I find far from acceptably comparable or appropriate, and (ii) the magnitude of a number of his adjustments, I found his testimony and his differing “other” approach to be at times helpful and informative.[39]

 

[151]   The Appellant’s only witnesses were Mr. Brennan, the Vice-President of Taxes of McKesson U.S. (and also on the Board of MIH and MIH2), Ms. Hooper of TDSI, and its experts Dr. Frisch and Mr. Reifsnyder. No one was called from McKesson Canada.

 

[152]   While Dr. Frisch suggested that he believed that arm’s length parties did non-recourse receivables purchase transactions on comparable terms and conditions and Mr. Reifsnyder suggested there was no reason why arm’s length persons could not do such transactions, no further evidence was given or called by the Appellant to substantiate that. Absent any supporting evidence, I am unable to give any weight or relevant consideration to the suggestions that there were such comparable arm’s length transactions being done.

 

[153]   Also, as Appellant’s counsel stressed and as is clear from the RSA and all of the evidence, one of the largest and most significant benefits to McKesson Canada of its receivable sales to MIH other than the cash purchase price rights,[40] was the transfer of the risk of loss on the receivables related to the Obligors’ credit and finances. According to the Appellant’s Objection to the reassessment, McKesson Canada effectively bought insurance against these risks by having them assumed by MIH as purchaser. There is credit risk or credit default insurance available in the market from arm’s length commercial players in the financial markets, both direct insurance and synthetic or derivative structured products.[41] However, the Court was given no evidence to be considered by it of the practical availability (or non-availability) or effectiveness, or pricing/costs for such. I found it somewhat surprising that neither side tendered any such evidence. I certainly think such evidence may have provided the Court with other helpful information with which to try to price the risk transference with information beyond McKesson Canada’s available historic loss and performance and projections for its receivables portfolio, and risk spreads in the public bond market.

 

7.       The Position of the Respondent

 

[154]   It is fair to say that the position of the Respondent shifted a fair bit through the trial. After the close of evidence and hearing the Appellant’s argument, the Crown conceded that the sale of any particular receivable under the RSA was non-recourse to McKesson Canada (the only exception being the limited right described above for MIH to require McKesson Canada to repurchase the receivable, initially at 75% of its face amount but ultimately only for what it was able to collect).

 

[155]   The Crown recognized that all of the other risk mitigation factors in the RSA in favour of MIH only allowed MIH to stop purchasing more receivables, or to recalculate the Loss Discount, on short notice in the event of expected or continued declining or deteriorating quality of the receivables, the Obligors, McKesson Canada or the McKesson Group, or material adverse change which could include market changes generally. There was no McKesson Canada covenant or assets supporting the collectibility of any receivable after it was transferred to MIH. This was recognized as fundamentally different than either a securitization or a secured loan. The Crown therefore resiled from its positions that would introduce structural reserves into the notional arm’s length RSA which were advocated, at least in the alternative, by two of its three experts. I think this was a wise decision.

[156]   In argument, the Respondent’s position with respect to the transfer pricing adjustment can be generally summarized by me as follows:

 

[157]   It is possible and appropriate to make adjustments to the Discount Rate pricing and terms and conditions in the RSA to reflect the terms and conditions that arm’s length parties would have agreed to, both by adjusting amounts and by adjusting the method or formula by which such amounts are directly determined or re-determined.

 

[158]   If the Court is concerned that the adjustment of a formula or criterion (e.g. fixed versus floating, yearly versus monthly et cetera) may not clearly be permitted under the language of paragraphs 247(2)(a) and (c), the OECD Guidelines’ commentary on “realistically available options/alternatives” would support a broader reading of which amounts’ quantum can be adjusted to reflect arm’s length terms and conditions. Such an adjustment would be a permitted adjustment under paragraphs 247(2)(a) and (c) and would not be a recharacterization of the RSA transaction itself described in 247(2)(b) and (d).

 

[159]   The appropriate adjustments are all to the risk factor mitigation issues and provisions described in the TDSI Report. The non-arm’s length terms and conditions can be identified, and the appropriate adjustments can be determined, having regard to the evidence from the experts in this trial and from TDSI (including Ms. Hooper’s evidence).

 

[160]   The Respondent called three expert witnesses to testify: Dr. Brian Becker, Mr. Joel Finard and Mr. Myron Glucksman. Each of these experts filed expert reports as well as rebuttal reports to the other side’s expert reports. Additional expert reports were filed by the Respondent with the Appellant’s consent and without viva voce testimony.

 

[161]   Dr. Becker first used a risk identification and assessment “build up approach” similar to those of TDSI and PwC generally following the structure of the RSA pricing provisions.

 

[162]   Dr. Becker also used an alternate comparable transaction approach focused entirely on McKesson Canada’s prior receivables factoring transaction with TD Factors. As described below, I do not accept that there is any helpful comparability between the RSA transactions and the prior TD Factors year end capital tax driven factoring.

[163]   Mr. Finard’s first approach was to use an “attribute analysis” identifying the risks inherent in the RSA from a credit and operations perspective and trying to quantify appropriate arm’s length pricing for each. As part of this approach, Mr. Finard sought to introduce a cash reserve to further protect against loss and thereby reduce the credit risk portion of the Discount Rate. While such a reserve would reduce the risk of loss to MIH on the transactions, it would do so by reducing the risk of loss on receivables after they were purchased by MIH which simply was not the deal in the RSA. As stated below, I am not sure that such is appropriate as a comparability matter, nor, without deciding it, am I sure that would be a permitted notional arm’s length term or condition to support an adjustment under paragraphs 247(2)(a) and (c).

 

[164]    Mr. Finard describes his second alternate approach as a “structured finance analysis”. In this approach Mr. Finard priced the credit risk portion of the Discount Rate by comparing McKesson Canada’s long-term historic loss experience on its receivables portfolio to rating agencies’ published loss rates by company debt rating, and used the public rating of companies whose public debt had a similar historic loss experience to identify the public debt market’s credit spread for a company with such a rating. I found this to be helpful and informative, recognizing, as with Mr. Reifsnyder’s approach, that public markets are only one particular market.[42]  To the extent there is comparability for the RSA transactions in the public debt market, Mr. Finard’s approach and identification of a rating somewhere between A and Baa makes much more sense than Mr. Reifsnyder’s approach using non-investment grade/high yield/junk bond rating status as a starting comparability point.

 

[165]   Lastly, Mr. Finard went through TDSI’s analysis, factor by factor.

 

[166]   The Respondent’s third expert witness, Mr. Glucksman, went through the credit risks identified in the TDSI reports and commented on their approach to pricing each of these as a component of the appropriate Discount Rate under the RSA. However, Mr. Glucksman’s approach dealt with the credit risk of the receivables’ Obligors differently than provided for in the RSA. He instead incorporated reserves into his arm’s length notional comparable transaction terms and conditions as a deferred purchase price which would be paid to the seller only as the transferred receivables performed. He then estimated the cost (operationally and time value of money et cetera) to the parties of maintaining such reserves. While it may be acceptable in paragraph 247(2)(a) and (c) adjustment of amounts reflecting arm’s length terms to look at pricing/cost of an alternative work-around or input for comparison purposes, the Glucksman reserves would have functioned very much like a reserve or hold-back or over-collateralization in a securitization transaction, or like the security on a secured loan, and leave the credit risk effectively with the seller, thereby making it effectively a recourse transaction. I have the same concerns with Mr. Glucksman’s reserves as I do with Mr. Finard’s reserve. I am able to fully dispose of the appeal without needing to decide to what extent, if any, the costs associated with such reserves, could help in a subparagraph 247(2)(a) and (c) analysis.

 

[167]   The Crown did not directly or indirectly raise any fair share or fiscal morality arguments that are currently trendy in international tax circles. It wisely stuck strictly to the tax fundamentals: the relevant provisions of the legislation and the evidence relevant thereto. Issues of fiscal morality and fair share are surely the realm of Parliament.

 

8.       The Witnesses, the Expert Reports and the PwC Report

 

          (a)  Mr. Brennan

 

[168]   Mr. Brennan testified on behalf of the Appellant, McKesson Canada. Mr. Brennan was the Vice-President of Taxes at McKesson U.S. throughout the relevant time.[43] He was also a director of MIH, the Luxembourg parent of McKesson Canada and of MIH2. He was neither an officer nor director of McKesson Canada. As mentioned above, no one from McKesson Canada testified at the hearing notwithstanding that it had a finance department and a large credit and collections department, each headed by accountants. No reason was given by the Appellant for the CFO and VP of Finance of McKesson Canada not testifying in this matter other than that his role was really only to provide financial information to TDSI to prepare its report. This was also the case for the head of Credit and Financial Services at McKesson Canada.[44]

[169]   Within the McKesson Group, corporate finance decisions were made by McKesson U.S. and dictated to McKesson Canada. All of the available evidence is that all of the relevant transactions were decided upon without any material input from McKesson Canada and solely by McKesson U.S., the 100% controlling shareholder of both parties to the RSA. The evidence is that these decisions were all made by the Treasury and Tax functions at McKesson U.S. in consultation with Blakes and TDSI.

 

[170]   The McKesson U.S. Tax department is responsible for tax compliance, (audits, filings and financial reporting) and for tax planning. The Tax department has been a rapidly growing group at McKesson U.S., having increased more than tenfold to 63 persons since Mr. Brennan took over the reigns in 2000.

 

[171]   Mr. Brennan described the $173,000,000 “double dip” ULC financing structure previously used since 1998 to finance McKesson Group’s Canadian holdings on a very tax efficient basis, effectively allowing both McKesson Canada and McKesson U.S. to deduct the same interest amount in both Canada and the U.S. for tax purposes. This double dip structure was paid out with a portion of the $460,000,000 received by McKesson Canada under the RSA in December 2002 from the initial receivables purchase.

 

[172]   Mr. Brennan also described the McKesson Group’s use of the Irish structure to hold all of McKesson U.S.’ non U.S. operations. He described this as very, very common Microsoft and Apple type international tax planning that relies upon the favourable tax treaties of countries such as the Netherlands and Luxembourg. It permits large multi-nationals like McKesson U.S. and its McKesson Group to amass large amounts of cash in Ireland by restructuring to avoid taxes that otherwise would have been paid in other countries.

 

[173]   MIH is a Luxembourg company that holds the shares of McKesson Canada and buys McKesson Canada’s receivables. In addition to Mr. Brennan, MIH’s board includes McKesson U.S.’ in house counsel as well as in house counsel of a McKesson Group UK company. In the year in question, MIH had a single employee. It does not appear from Mr. Brennan’s evidence that MIH had a second employee or leased any premises in Luxembourg until some time after the years in question.

 

[174]   Mr. Brennan testified that the McKesson Canada DSO was in the 30 to 31 day range at the time of the RSA and remained at around 30 days.

[175]   According to Mr. Brennan, MIH would look at the actual aging of McKesson Canada’s receivables regularly. MIH would concentrate on the Designated Obligors because “that’s really where risk of this company is at”. MIH would look at defaults of receivables and would decide whether or not it would exercise its rights under the RSA to sell them back to McKesson Canada as described above. It is not clear why MIH would not always put defaulted receivables to McKesson Canada; it does not appear to be consistent with arm’s length behaviour.

 

[176]   Mr. Brennan testified that at the time of the RSA, McKesson Canada had access to US$100,000,000 to $150,000,000 of a US$550,000,000 McKesson U.S. one year facility. McKesson U.S. also had access to another US$550,000,000 standard credit facility which apparently could have been used by McKesson U.S. to finance McKesson Canada. McKesson U.S. also had access to a one year securitization program, which renewed annually, and could also have been used to finance McKesson Canada. These were not so used because it would not make sense and would be inefficient to borrow money at interest from third parties when all this surplus cash had been amassed by McKesson Group in Ireland for this very purpose. Later in his testimony, Mr. Brennan added that withholding tax on interest would also cause further inefficiency.

 

[177]   Mr. Brennan’s evidence emphasized the significantly increased cash flow to McKesson Canada from the RSA. As noted previously however, since the McKesson Canada receivables pool had a DSO in the 30 day range, and the RSA settlement period was also about 28 days, the cash flow impact would only arise upon and from the initial December 2012 sale of the pool partway through a settlement period and only in respect of a portion of that amount. Otherwise, McKesson Canada largely continued to receive cash on a similar schedule throughout each roughly 30 day period thereafter, that is, as its customers paid their bills. This largely one-time pick up would not be insignificant, and would effectively be permanent for the remainder of the term.

 

[178]   Mr. Brennan described how he did a short, quick calculation of the net tax benefit to McKesson Group of implementing the RSA and Servicing Agreement. The hand written calculation he had prepared for McKesson U.S.’ CFO was put into evidence. It showed that there would be a Canadian tax reduction of about US$4.5 million in the short year ending March 2003, Luxembourg tax of US$29,000 for the year ending March 2003, and costs associated with maintaining MIH of US$300,000 of bank charges and US$35,000 for accounting fees. The net tax benefit to McKesson Group was over US$4.1 million in the three and one-half months remaining of McKesson Canada’s 2003 year and about US$15,000,000 annually thereafter. These numbers assumed only US$250,000,000 of McKesson Canada receivables were generated monthly. The net Canadian tax savings would be corresponding larger since C$460,000,000 of receivables were sold in December 2002.[45] It appears this analysis was not shared in any way with the CFO of McKesson Canada.[46]

 

[179]   Mr. Brennan did not read the TDSI Report before closing the RSA transactions. He is not sure he saw the TDSI engagement letter. He simply got their Discount Rate number and used that to do his tax savings calculations. Mr. Brennan testified that McKesson US had looked at financing alternatives internally but just did not ever write them down. He said there was no internal review done of TDSI’s numbers, nor did McKesson Group do any sensitivity analysis to assess the reasonableness of any alternative discount rate.

 

[180]   Mr. Brennan could not recall why the RSA had a five-year term. He did say it was on the advice of TDSI. This is not supported by the TDSI Report or Mrs. Hooper’s evidence. He could no longer recall why the Servicing Agreement set an annual fee for the servicer of $9,600,000.

 

[181]   Mr. Brennan spoke several other times in his testimony about the role of TDSI in structuring the transaction. He said TDSI reviewed McKesson Canada’s credit and collection policies. Ms. Hooper testified afterwards that it had not; nothing in the TDSI Report suggests it did. He also described TDSI as having been retained to tell McKesson U.S. what the RSA Discount Rate should be as they did not have the banking expertise to do this in-house. Not only does this appear to be an overstatement of Ms. Hooper’s testimony and the language of the TDSI Report, it also appears odd that McKesson U.S. not having the relevant expertise, turned to TDSI who stressed in their reports and in Ms. Hooper’s testimony that the RSA was also outside its expertise. Further, Mr. Brennan said it was TDSI that recommended the RSA be a $900,000,000 facility even though that significantly exceeded the amount of receivables at the time. This is also unsupported by the TDSI reports and inconsistent with Ms. Hooper’s testimony. I accept Ms. Hooper’s testimony in all of these regards.[47] She was much less self-interested and displayed more candour in her overall testimony. Her version was consistent with the TDSI reports.

 

[182]   I conclude that Mr. Brennan’s memory on these key aspects of these transactions has proven, for whatever reason, faulty or lost. This does not make my task any easier, especially since he was the only witness from the McKesson Group.

 

[183]   Mr. Morgan of Horst Frisch advised Mr. Brennan in the summer of 2002 that such an RSA transaction would be so unique that he recommended that an expert in the area should be retained for a transfer pricing study. This appears to have resulted in Ms. Hooper at TDSI’s role and the TDSI reports. However, Mr. Brennan testified in cross examination that neither McKesson U.S., nor McKesson Canada sought a formal transfer pricing study.

 

 (b)  Ms. Hooper and the TDSI Reports

 

[184]   I have already fully described the TDSI reports.

 

[185]   Ms. Hooper described the original TDSI retainer as a one-off situation, that she had never done something like this previously. She had never previously priced nor been involved in pricing risks in receivables transactions because in securitizations risks are avoided not purchased. She said the same was true of the TDSI Servicing Agreement report, that she had no prior experience doing such an analysis. She also acknowledged she had no experience with traditional factoring of receivables in a non-securitization environment.

 

[186]   In her testimony, Ms. Hooper identified the material transaction risks as (i) dilutions, including prompt payment discounts (ii) losses – primarily credit losses and (iii) servicing risk, that the servicer collects but does not remit to the buyer.[48]

 

[187]   Ms. Hooper clarified in evidence that, while the TDSI Report says that where possible it looked at pricing of comparable risks in the market, TDSI did not really do that beyond obtaining credit spread numbers from their bond traders. It appeared from the TDSI reports and Ms. Hooper’s overall testimony that this may have been because of her being a securitizations expert entirely unfamiliar with the pricing of such risks in the market. To the extent that is the case, one can assume that the McKesson Group and its advisors were aware of that.

 

[188]   She explained that the DSOs used in the TDSI Report were computed by TDSI from historical data received from McKesson Canada. TDSI’s DSO number simply divides the receivables balance at the end of a McKesson Canada Accounting Period by the sales in that period, and multiplies that by 30.[49] DSO numbers serve as a proxy for accounting purposes for how long receivables can be expected to take before payment in full. However, a DSO number simply compares the receivables amount at the end of the period with the sales during that period. Thus, it does not actually reflect the time it takes for any particular receivable, nor the receivables pool on average, to be paid. The DSO will by definition increase if sales go down in a period and will decrease if sales go up in a period, even if nothing in fact changes with respect to payments on receivables in that period. No one questions that it is nonetheless an appropriate and satisfactory proxy.

 

[189]   Ms. Hooper also explained that the historic average DSO would not have been a good proxy for the initial receivables pool transferred in December 2002, as these were not transferred on the day receivables originated as would be the case thereafter under the RSA. To address this, TDSI instead estimated it would take half as long to collect the initial mature receivables pool and spread the “missing” 16 days over the five-year term. Neither the TDSI Report nor Ms. Hooper’s evidence acknowledge or discuss that this effectively overstates the Discount Rate in MIH’s favour and that the underpayment is only recouped by McKesson Canada over five years without interest. It can also be noted this gives McKesson Group a five year timing benefit in its Canadian income taxes.

 

[190]   With respect to the 20% buffer added by TDSI to the historic prompt payment discount multi-year average numbers, Ms. Hooper could not recall or explain how that 20% number was arrived at. She speculated that TDSI may have just looked at the possibility that 100% of Obligors would exercise their prompt payment discount rights. Again, the fact an increase in prompt payment could be expected to have a significant impact on DSO calculations was overlooked.

 

[191]   Ms. Hooper confirmed in her testimony that she did not have any experience calculating loss discounts in her securitizations area of expertise either. This is because, in securitizations, credit loss risk is addressed through structuring credit enhancements and other risk mitigators, not through price.

 

[192]   With respect to the TDSI Report’s statement that the historic information provided on write-offs to sales in considering the Loss Discount did not cover a full Canadian economic cycle, it does not appear from her testimony that TDSI asked McKesson Group for further data nor was told it was unavailable.

 

[193]   Ms. Hooper had difficulty, even in direct examination, giving a satisfactory, responsive, logical or complete explanation of the interest discount portion of the Discount Spread relating to MIH’s cost of funds given it was a thinly capitalized company. With respect to MIH’s cost of funds interest discount, she testified that TDSI had simply assumed MIH was 100% debt financed even though their information was that it was thinly capitalized. She has no recollection the McKesson Group ever told her of the source or terms of MIH’s funds used to purchase the McKesson Canada receivables.

 

[194]   Ms. Hooper added that the delinquent portfolio performance trigger serves as an early warning system. Typically with receivables one will see delinquencies increase in advance of seeing losses increase. For this reason, she explained one wants the trigger rate to permit termination of the agreement early enough for there to not be material losses. She clearly understood that when the transaction could be terminated would have a fairly significant affect on the overall risk that was being transferred. According to Ms. Hooper, portfolio performance triggers, delinquency and default rate triggers, were designed to limit the ultimate losses to the purchaser by ceasing the acquisition of new receivables that might not be expected to perform as well as receivables originated previously.

 

[195]   Ms. Hooper explained in her testimony that in a typical securitization transaction one uses historical performance to try to get comfortable with how performance might be expected to unfold in the future.

 

[196]   Ms. Hooper said in her testimony with respect to the servicing discount that, while the likelihood of the delinquency or default termination ratios being triggered can be assessed relative to historical performance, it is more difficult to quantify in any reliable fashion the likelihood of other termination events occurring. She could not explain how TDSI got from the 9.41% credit rating migration risk number of a potential downgrade of McKesson U.S.’s credit rating (which would give rise to a termination event) to the report’s 25% total chance of a termination event occurring. Nor did she explain why the TDSI Report assumed that MIH would exercise its rights to terminate 100% of the time a termination event occurred no matter what the circumstances.

 

[197]   Ms. Hooper believed TDSI continued to be involved, after the two TDSI reports in evidence, in assisting with the annual review of the calculation of the RSA Discount Rate required by the RSA.

 

[198]   Ms. Hooper could not provide a very thorough or satisfactory answer as to why the TDSI Report, when assessing the accrued rebates dilution risk, assumed that 100% of McKesson Canada’s customers would set off their full accrued rebates. She could not explain how this was reasonable. She could only repeat that TDSI chose the maximum number to be conservative as mentioned in its report. She added that the accrued rebate dilutions discount did not make up a very significant portion of TDSI’s overall Discount Rate and, perhaps if it had been more significant, they might have considered whether average rebate numbers would have been more appropriate.

 

[199]   The TDSI Report was qualified with soft opinion language. That indicates to me that it was written by TDSI and understood by McKesson Group to be primarily advocacy. The shortcomings in the report and in her testimony described above served to confirm that.

 

 (c)  The PricewaterhouseCoopers Report

 

[200]   McKesson Canada and its advisors had PricewaterhouseCoopers (“PwC”) prepare a transfer pricing report relating to the RSA in December 2005. The PwC Report was used to respond to CRA’s review of the RSA transactions.

 

[201]   PwC followed much the same approach as TDSI but without any supportive testimony at trial. I can give it little weight except to the extent it supports the use of the TDSI approach. I do find some of the data/information relied on or referred to in the PwC Report to be corroborative of, or complementary to, some of the numbers/ranges/approaches/observations used by others. I have identified below where I considered any such information useful.

 

[202]   It is clear from the PwC Report that it considered factoring receivables to a commercial financial market player could be a potential comparable. It said that PwC had looked at the range of factoring yields of third party factoring companies to support its view that there should properly be a range for an arm’s length Discount Rate. Later, it looked to a range of participants in the North American factoring market to support its view that, as a highly leveraged entity, MIH’s cost of funds should be borne by McKesson Canada as discussed further below. Yet, the PwC Report did not develop the concept of arm’s length factoring and the pricing or other terms and conditions thereof as a comparable transaction to the RSA, nor did it address why that would not be possible or appropriate. This is a significant shortcoming and causes one to think that the PwC Report was primarily a piece of advocacy work, perhaps largely made as instructed.

 

[203]   The PwC Report notes that the RSA fixes a DSO number instead of using a dynamic floating DSO. As part of this, PwC looks to the historical actual performance of McKesson Canada’s pool of receivables to inform its views. The report specifically notes that, with respect to the DSO number in the RSA, the interests and exposure of both McKesson Canada and MIH need to be considered and balanced. However, PwC’s conclusions are then limited to MIH’s potential exposure to adverse actual performance of the DSO and PwC has “taken the position” and “assumed” a 20% cushion should pad the fixed DSO number to incorporate a 38 day DSO. PwC did not address in any way the flipside of the coin – McKesson Canada’s risk in fixing the DSO at any number for the term that its DSO actually continued to improve. Nor did PwC address why incorporating a variable dynamic DSO which would almost fully protect both MIH and McKesson Canada would not be more appropriate than simply increasing the amount of the already fixed[50] DSO.

 

[204]   By adding a 20% cushion to the fixed DSO, the PwC approach makes MIH favourable upwards adjustments to all of the factors comprising the RSA Discount Rate that are determined by reference to annual rates, including the RSA’s risk free 30 day CDOR baseline Yield Rate, by 20%.

 

[205]   It can be noted that PwC looked at one year, not five year, securities when looking at the Obligor Loss Discount factor. This suggests PwC did not equate the RSA’s five-year term with the risk associated with the five-year term financings for credit risk purposes.

 

[206]   The PwC Report refers to debt issues with a below investment grade rating as junk bonds.

 

[207]   The PwC Report also looked at the actual historic loss performance of the McKesson Canada receivables pool as part of its analysis. However, it did not do so as a starting point to projecting the credit loss risk associated with the pool. It assessed credit loss risk by first assigning a Loss Discount attributable to the Designated Obligors equal to the credit spread on public bonds issued by what they believed were similarly rated corporate issuers in the public bond market. This they then adjusted for their cushioned DSO. The PwC Report did not try to reconcile its chosen approach to the actual known performance of the receivables pool.

 

[208]   However, in dealing with the other Obligors credit risk, PwC did analyze the receivables pools’ historic performance, but only to demonstrate that the other Obligors’ receivables historically had losses exceeding that of the Designated Obligors. PwC then used this to justify notching all of the others Obligors’ credit risk down two grades below the public debt rating PwC had assigned to the Designated Obligors (before also adjusting it for their cushioned DSO). This picking and choosing, mix and match as it suits approach to the relevance of the actual performance of the receivables pool makes for transparently poor advocacy, and even more questionable valuation opinions.

 

[209]   The TDSI Report had assumed a 25% chance that a replacement servicer would need to be appointed when it addressed the servicing discount portion of the Discount Spread. TDSI relied upon Moody’s one year number for a potential downgrade of McKesson U.S. occurring (which would be a termination event). PwC raised this to a 40% probability relying upon Standard & Poor’s five year downgrade number.[51] As with the TDSI Report, the PwC Report did not address why a change of servicer would be a requisite necessity upon the occurrence of such a termination event. The RSA and Servicing Agreement do not terminate McKesson Canada’s appointment as servicer upon a termination event. They merely give MIH the right to appoint a new servicer if it chooses. Given that (i) the particular termination event of a credit rating downgrade of McKesson U.S., which both TDSI and PwC rely upon in support of their 25% and 40% numbers, would not necessarily involve any change to McKesson Canada’s ability to service its receivables, and (ii) given that MIH had the right in the agreements to end commingling and require segregation of funds, thereby effectively removing McKesson Canada credit risk, these probability numbers remain inadequately explained, questionable and, to my mind, unreliable advocacy or posturing.

 

[210]   It can be noted that PwC’s replacement servicer fees were in the range of 0.79% to 1.23% of the face amount of receivables. The midpoint of this range is 1.01%.

 

[211]   It should also be noted that PwC looked at and relied upon McKesson Canada’s four year historic prompt payment discount when considering the prompt payment dilutions discount component of the Discount Spread. While TDSI added an unexplained 20% buffer to the historic actual 0.5% experience of McKesson Canada’s receivables pool, PwC “assumed” that an arm’s length purchaser would want a 5% cushion. This assumption was similarly unexplained and, tellingly, is an upwards adjustment that only addresses MIH’s risk that more Obligors pay more quickly. It does not address McKesson Canada’s risk that fewer Obligors take advantage of their prompt payment discount. Nor did PwC account for the impact that more early payments from Obligors would have upon the DSO, and which would also benefit only MIH.

 

[212]   At this point in its report, PwC summarized and arrived at a range of arm’s length Discount Rates for the RSA of 1.3179 to 1.4823, still quite a distance from the rate agreed to in the RSA between McKesson Canada and its parent MIH, and that in the TDSI Report – indeed much closer to that used by CRA in the reassessment.

 

[213]   However, the PwC Report then went on to add a further factor to that range that is essentially TDSI’s interest discount, which effectively imposes on McKesson Canada the full amount of an assumed cost of funds of MIH.[52] PwC does not address the fact that MIH is borrowing all of the money at a cost of funds determined by reference to the RSA Discount Rate. PwC assigns a credit spread based upon public debt issuers which are 90% capitalized with debt, which are below investment grade rated companies by Standard & Poor’s. This factor alone then drives the PwC range for an arm’s length discount rate from the above numbers to a range of 1.9698 to 2.2646.

 

[214]   PwC’s sole expressed reason for thinking that MIH’s cost of funds would be agreed to be borne by a notional arm’s length McKesson Canada was because five-year terms are not generally agreed to by large participants in the factoring market and that this premium to the Discount Rate could therefore be demanded by MIH. PwC ignored the presence and availability of large well-capitalized financial institutions in the factoring market that it had already described in its report. PwC looked instead at their smaller and much weaker, poorly capitalized competitors as though McKesson Canada was driven by market forces to use one of these latter players. McKesson Canada probably did have to use MIH, but only because of McKesson Group’s control, not because of any market, economic or financial reasons that are in evidence. There was no evidence that McKesson Canada or McKesson Group was even interested in considering factoring its receivables to any arm’s length financial institution player in factoring markets, presumably because profits would then have left the McKesson Group.

 

[215]   It should also be noted that PwC could only get this factor so high by assuming that MIH, the poorly capitalized player, should also nonetheless command the level of premium yields, net of its high cost of funds, realized by the well-capitalized large players described in the PwC Report. The PwC Report numbers were not reflective of the lesser yields earned by poorly capitalized players as they themselves described the market. More unsupported, selective picking and choosing.

(d)  The Frisch Expert Report

 

[216]   Dr. Frisch has a PhD in Economics from Harvard University. He is an expert in transfer pricing methodologies. In his opinion, none of the named, recognized OECD Guidelines’ methodologies could be applied in a reliable manner to the RSA transactions. He was of the opinion that an “other methodology” needed to be developed, as contemplated by paragraph 2.9 of the 2010 OECD Guidelines. He volunteered that a transfer pricing economist such as he did not have the expertise to develop such an “other method”.[53]

 

[217]   Based upon what evidence the parties chose to put before me, I agree with Dr. Frisch’s opinion that the named OECD methods are not appropriate to rely upon in this case, and that an “other method” should be used.

 

[218]   I am surprised that Dr. Frisch thought that developing or opining on an “other method” was outside the expertise of expert transfer pricing economists. Clearly, the Respondent’s expert, Dr. Becker, did not share the same reticence.

 

(e)  The Reifsnyder Expert Report

 

[219]   The Reifsnyder Report, after downward corrections and amendments made in the course of trial, opines that arm’s length parties “would have agreed to” a discount rate to be applied to the face amount of the receivables in the range of 1.7326% to 2.4360%.

 

[220]   Mr. Reifsnyder’s approach was to consider, estimate and add up four factors:

 

(i) Factors One and Two – Servicing Fees and Prompt Payment Discounts

 

[221]   Mr. Reifsnyder began by determining the anticipated cash flows to the purchaser in the transactions. In addition to the cash flows expected by the purchaser from the face amount of the receivables, he identified his first factor as the fixed monthly outbound servicing fee payable under the Servicing Agreement, and his second factor as the cash flow reductions that could be expected from prompt payment discounts enjoyed by McKesson Canada’s customers. With respect to prompt payment discounts he used McKesson Canada’s historic three-year average prompt payment discount numbers. He did not see the need to be concerned with the risk of change since, in his experience, early payment discounts were not usually volatile in respect of trade receivables. In his experience, it was not uncommon for prompt payment discounts to be borne by a buyer in a receivables purchase transaction.[54] He does not identify any cash flow reduction potential from McKesson Canada’s customers setting-off rebate entitlements against the receivables.

 

          (ii)     Factor Three – Credit Risk

 

[222]   Mr. Reifsnyder then developed a third factor to reflect the credit risk on the receivables portfolio. This third factor focused on the credit risk associated with the Obligors, not McKesson Canada.

 

[223]   His report accepts the Yield Rate component of the Discount Rate in the RSA, being the CDOR 30 day BA rate as the appropriate baseline risk-free rate of return to which an appropriate spread should be added.

 

[224]   He next determined the appropriate credit risk spread by starting with the spread on a particular Canadian bond index fund that he viewed as sufficiently comparable, provided a number of upwards and downwards adjustments were made to it, to reflect what he thought an arm’s length potential purchaser would seek if presented with the RSA.

 

[225]   Recognizing that the RSA was a five-year agreement, dealing with Canadian dollar denominated receivables, with some large concentrations in the pool with the Designated Obligors, with some security but no credit enhancements, and with the majority by value of Obligors not being rated, Mr. Reifsnyder selected a particular Canadian high yield bond fund index. He described a bond fund index as an aggregation of financial assets that share common and similar characteristics.

 

[226]   The credit spread on his chosen Merrill Lynch Canadian Dollar High Yield Index for the six reported months prior to the RSA reflected a spread of over 13% per annum (and almost 14% per annum for the last reported month) above the rates of return on risk-free Canadian Treasury obligations. As of November 2002, that is when the RSA was being priced, this index was composed of 26 bonds issued by 18 issuers, each of which was rated below investment grade – that is the index was comprised of high yield bonds also referred to as junk bonds. The minimum bond amount was $50,000,000 and the minimum term to maturity was one year. A number of the individual bonds making up this index were trading at discounts in excess of 50% and one was discounted 98%. A number had effective yields in excess of 30%, the highest being 499%.

 

[227]   For the reasons detailed below, I do not accept that the Merrill Lynch Canadian Dollar High Yield Index was an appropriate starting point.

 

[228]   The Reifsnyder Report then made a number of upward and downward adjustments to the High Yield Bond Fund Index spread to reflect the particulars of the McKesson Canada receivables pool as compared to the bonds making up the bond fund index.

 

[229]   The first adjustment was to reflect the greater single industry concentration and correlation risk in the McKesson Canada receivables pool. This was an upwards adjustment to the bond fund index spread of 15% to 20%.[55]

 

[230]   The second adjustment was to reflect the greater proportion of small business Obligors in McKesson Canada’s receivables pool compared with the small number of public companies in the bond fund index. A further upwards adjustment of 10% to 15% was made to the spread. There was no discussion of whether small going concern businesses could really be considered riskier than larger ones that, in the case of this index, had to be rated junk status to be in the index.

 

[231]   The third adjustment reflected that a portion of the Obligors in the McKesson Canada receivables pool were large investment grade rated companies, and others were hospitals. Mr. Reifsnyder made a McKesson Canada favourable downwards adjustment to the bond index spread of 15%, roughly the approximate makeup of such Obligors in the pool when the RSA was entered into.

 

[232]   The fourth adjustment reflected the fact that the Obligors in the pool could change throughout and that there was no limit in the RSA on the concentration risk of the Designated Obligors. In his opinion, this contrasted unfavourably with the bond fund index and warranted a further upwards adjustment of 20% to 40%.

 

[233]   The fifth adjustment of 5% to 15% upwards reflected the fact that the adjusted duration to payment or maturity of the bonds in the index was just under three years whereas “[s]ince the outstanding balance on the facility under the RSA could be as a high as the facility limit [$900,000,000] up to the Termination Date in 2007, the duration of the [McKesson Canada] portfolio is considered to be five years.” This is a non-sequitur; there is simply no obvious or described causal relationship of the facility limit to its term. Also, as mentioned previously, the term of the RSA as an agreement and contractual obligation should not be confused or conflated with five-year term money. Neither Mr. Reifsnyder nor any other of the Appellant’s witnesses explained why a notional arm’s length McKesson Canada would want a $900,000,000 facility in its particular business circumstances, present or projected.

 

[234]   A favourable downwards sixth adjustment of 25% to 40% was also made to reflect his view that conventional wisdom is that short-term trade obligations like the receivables are less risky than long-term loans like the bonds making up the index. There was no attempt to reconcile this thought with his previous adjustment’s position.

 

[235]   His seventh and final adjustment was a further upwards adjustment of 5% to 10% to reflect his view that a trade supplier’s security on inventory compares unfavourably to pledges of assets by bond issuers in the index. He does not address the probability of senior ranking security on inventory as compared with the improbability of security or probability of numerous prior secured creditors of junk bond issuers. The evidence is that McKesson Canada’s credit policies did require specific security in appropriate cases and could even require seats on a customer’s Board of Directors.

 

[236]   These seven adjustments in total were between 95% and up to as high as 155% of the starting point of his chosen bond fund index. Two adjustments were up to 40% each. While his report gave some explanation for some of the numbers he assigned, and some more was given in his testimony, these percentages were largely based upon his judgment and experience – though he admitted to having no experience pricing a trade receivables financing transaction off a high yield bond index.[56] None of this prevented him from opining on an arm’s length discount rate range to five significant digits, surely an extreme example in false precision (to use a phrase from his own report).

 

[237]   The Reifsnyder approach resulted in a credit risk component of the Discount Rate in the range of 0.85% to 1.6%. This is a broad range and is up to almost 40 times higher than the 0.43% historic loss performance of McKesson Canada’s receivables pool, and more than six times the 0.25% Loss Ratio termination event trigger in the RSA.

 

(iii)    Factor Four - $900MM Commitment to Finance

 

[238]   Lastly, Mr. Reifsnyder added a factor to his arm’s length Discount Rate for the RSA to reflect that the RSA committed MIH as purchaser to purchase up to $900,000,000 of eligible receivables over the five-year term. In his opinion, this warranted a standby charge/commitment fee type component in the Discount Rate to reflect the unused portion of the facility at any time during the five years in order to compensate MIH as purchaser for McKesson Canada’s option to obtain increased funding up to the full $900,000,000.

 

[239]   His report notes that the bank market annual fee to a low investment grade rated borrower for five years has been “in the range of .00375% to .0050%”. He opines that an appropriate rate for the RSA commitment would be “at least 1%” given the risky nature of the receivables pool. His written report’s “at least 1%” for the RSA is 200 to 300 times greater than the report’s “range of .00375% to .0050%” in bank markets for low investment grade borrowers. One might therefore have expected greater explanation in the report than just more risky.[57]

 

[240]   His report does not address the fact that MIH is a very thinly capitalized, non-bank, non-financial institution that is not regulated as to its capital adequacy, that otherwise has zero presence in the financial sector or financial markets, and whose principal asset is its somewhat illiquid shares of McKesson Canada itself.

 

[241]   His report does not address why a notional arm’s length McKesson Canada would pay for a standby facility that was almost double the amount needed.

 

[242]   There was no evidence to suggest that an arm’s length counterparty to an RSA done in the commercial markets with a member of the McKesson Group would be such a non-investment grade entity.

 

[243]   Overall, I do not accept that there is any merit or value on the facts of this case to using Mr. Reifsnyder’s methodology or his assigned inputs thereto to help determining an appropriate arm’s length Discount Rate for the RSA. That is not to say, however, that I reject his evidence completely.

 

[244]   In short, I reject his chosen high yield bond fund index as an appropriate starting point because of the make-up of that index. The scope and amount of the adjustments he needs to make to account for its differences from McKesson Canada’s receivables pool further confirms that it is not sufficiently comparable to serve as an appropriate starting point. In my view, the appropriate analysis places no weight on his methodology and approach as I find it not to be an appropriate “other method” to determine an arm’s length RSA Discount Rate. I do not therefore place any weight on his opinion of the appropriate range for an arm’s length Discount Rate.

 

[245]   My specific detailed concerns with Mr. Reifsnyder’s approach and opinion are as follows:

 

(a)     I do not have sufficient evidence to accept that there is any significant degree of comparability of such a bond fund index to a pool of inventory receivables owed by going concern customers to a significant going concern supplier. Mr. Reifsnyder had never priced a trade receivables transaction off any high yield bond fund index, nor could he say he had ever heard of or seen others do so. He was not ever sure he had ever before looked at the Merrill Lynch high yield bond fund indexes relied upon in his report. Similarly, none of the other expert witnesses asked could say they had ever heard of it being done. Mr. Reifsnyder did not testify that he had ever himself priced, or seen another price, any structured finance transaction (other than perhaps a bond, bond fund, or a bond fund index) off a bond fund index. The Respondent’s experts who were asked said they had never heard of one.

(b)     The high yield index is made up of bond issues, corporations that have raised money in the public markets generally for general corporate purposes. In contrast, the Obligors’ receivables in McKesson Canada’s receivables pools receivables are owing to one of their key suppliers for the inventory they need to remain in business as a going concern.

 

(c)      Ms. Hooper of TDSI testified that in Canada there was not at the time, and especially in 2003, a very liquid market for sub-investment grade bonds.

 

(d)     There are only a handful of issuers in the high yield index. McKesson Canada has many multiples that number of Obligors.

 

(e)      Regardless of the quality of the issuer in the index when they first issued their bonds, by definition their bonds are only added to the high yield index if they are of junk status. In extreme contrast, McKesson Canada’s Obligors at the time the RSA was entered into had regularly paid their receivables in full about 99.96% of the time. While there was actually a risk this default rate could change, there was no evident expectation that it would. In December 2002, these would be two drastically different snapshots virtually incapable of being more different absent actual default.

 

(f)      McKesson Canada’s Obligors were monitored and rated by McKesson Canada’s Credit and Collections department, and the terms of credit, if any at all, offered to them could be changed by McKesson Canada at a moment’s notice. In contrast, the bonds reflected in the chosen high yield index were for a term with either a lump sum payment upon maturity years later, or perhaps a series of annual payments through maturity. That is, a bond holder paid its money, took its chances and had to wait and see or sell. According to Moody’s, the historic credit loss on five year junk bonds is in the range of 27%. This is almost 700 times the 0.043% historic loss on the McKesson Canada receivables pool.

 

(g)     The individual discounts on particular issuers’ bonds in the index were as high as 98%. The five year bond issues mentioned above that traded at less than fifty cents on the dollar and had effective yields in excess of 30% comprised 8.5% of the index in November 2002. There is absolutely no evidence to even suggest that a single one of McKesson Canada’s Obligors, including the Designated Obligor whose financial situation was detailed during the trial, was in even remotely comparable financial circumstances.[58]

 

 (h)    The high yield bond fund index spreads fluctuated very significantly. In the six months prior to the RSA, the index spread’s performance had fluctuated by almost 50%. In contrast, in the same period, the McKesson Canada receivables pool’s performance remained at a roughly similar .04% write-offs to sales number and a DSO within the 30 day range. Further, there was no evidence given to Mr. Reifsnyder or to the Court of any significant change in the public ratings of those of McKesson Canada’s customers that were rated, nor to the credit ratings or evaluations of all of its customers by its own Credit and Collections department. This volatility and fluctuation in the range of discount spreads on the high yield index was such that had the bond fund index spread for a different month been used as the starting point for Mr. Reifsnyder’s approach, his approach’s arm’s length Discount Rate would have been materially different even though there is no evidence to suggest the quality of McKesson Canada’s receivables pools was materially different throughout the period. Mr. Reifsnyder did not try to relate this index’s OAS volatility to fluctuations in interest rates or credit spreads during the same period.[59] [60]

 

(i)      the credit risk spreads on the bond fund indexes are properly known as Option Adjusted Spreads or OAS. The OAS is widely viewed as the credit risk premium over a risk-free rate that applies to issuers in a similar class. OAS are not simple mathematical calculations. They are calculated using quantitative models and quantitative assumptions. Mr. Reifsnyder has never calculated an OAS and could not at all describe the models or assumptions used either generally or in the case of his chosen index.

 

(j)      There was no evidence that any of McKesson Canada’s Obligors that had public ratings, or those whose parent company was rated, had ever had their rating downgraded at all, much less to anything comparable to high yield junk status.

 

(k)     Mr. Reifsnyder acknowledged that in the case of McKesson Canada’s  rated Obligors, using their individual and particular bond rating credit spreads could be more accurate than his approach – he just did not know and did not explain why he did not consider that even though the results would be very different.

 

(l)      Notwithstanding the fluctuating volatile spread of the high yield index OAS in the prevailing market at the time, Mr. Reifsnyder’s approach used a fixed credit risk discount spread throughout the five-year RSA term and did not use a dynamic floating spread for the index to adjust for or reflect this underlying volatility. His position was that while a dynamic approach to credit losses would be more intuitive, they would constitute the use of hindsight.  I could not disagree more. A dynamic approach during the term of an agreement and provided for in an agreement is either not the use of hindsight as that term is used in circumstances such as these, or is at least not an inappropriate use of hindsight. He did acknowledge that there is a continuing dynamic adjustment in credit risk assessments in securitizations to reflect the most current performance throughout the term of the transactions.

 

(m)    The adjustments to the high yield index proposed by Mr. Reifsnyder to make it comparable to McKesson Canada’s receivables pool and the RSA were of orders of magnitude in the aggregate (up to 155%), and at least some individually (two up to 40%) which belie any starting point suitability for pricing comparability purposes.

 

(n)     As noted above, the explanations and premises for several of his proposed adjustments to the high yield fund index were inadequately supported and/or insufficiently explained.

(o)     Mr. Reifsnyder’s “at least 1%” stand-by commitment fee for the undrawn portion of the $900,000,000 facility limit under the RSA can not be supported or accepted at all (i) absent any explanation for his 20,000% to 30,000% adjustment to his 0.00375% to 0.0050% range data point for low investment grade borrowers in bank markets, (ii) absent any understanding of why the $900,000,000 facility limit would be of value to McKesson Canada throughout its term and that was reasonably supported in the evidence, and (iii) absent any adjustment for the fact that MIH was not a bank but an unregulated private company with virtually no capital or financial market presence, nor any explanation for the absence of need for any such adjustment. This component of Mr. Reifsnyder’s opinion of an arm’s length RSA discount rate is not only wholly rejected, it raises further concerns with his overall report.

 

(p)     Notwithstanding his extensive knowledge, experience and presentations on credit insurance for structured finance transactions, and his reference to its availability in his testimony, Mr. Reifsnyder seemingly never considered the cost of insuring the receivables in his pricing approach, nor to test the results of his approach. Nor did he explain why he did not do so.

 

(q)     While Mr. Reifsnyder maintained that markets should not price risk off historical performance, he could not say that markets did not. His view is at odds with that of TDSI and Ms. Hooper who refer several times to their need to consider historical performance in terms of defaults, delinquencies, losses, dilutions, payment times and reserves. It is also at odds with the rating agencies’ publications on the significance of historic performance in trade receivables financing structures and the evidence of other experts who testified. Historical delinquency and write-off performance is generally considered to be the best indicator of portfolio credit quality in trade receivables transactions. Mr. Reifsnyder acknowledged that rating agencies look at five years of historic performance data when reviewing a portfolio of financial assets. The Reifsnyder Report and testimony try to downplay the relevance of the historic performance of the McKesson Canada receivables pool in favour of the actual, then current credit spreads of its chosen high yield index. The irony is that the high yield index credit spread is itself a reflection in part of its historic performance and that of the issues and issuers it comprises. 

 

(r)      Mr. Reifsnyder came across as in large measure a partisan advocate quick to point out the specks in the Respondent’s expert reports, and downplaying, if not refusing to acknowledge, the weak points in his own.

 

[246]   Overall I can say that never have I seen so much time and effort by an Appellant to put forward such an untenable position so strongly and seriously. This had all the appearances of alchemy in reverse. One could only assume that the Appellant knew full well the weaknesses of the TDSI Report and this was the best method it could use to support the Discount Rate used by the McKesson Group in the RSA.

 

  (f)  The Becker Expert Report

 

        Build up approach:

 

[247]   The first valuation approach in the Becker Report relied upon by the Respondent, which was described as a build up approach, essentially considered the same factors and issues as the TDSI Report to arrive at a Discount Rate that reflects both the time value of money and the potential risk of not collecting the full face value of all of the transferred receivables.

 

[248]   The Becker Report accepts the 30 day CDOR rate as the appropriate risk-free rate, noting it is common to use government obligations of similar maturities as a bench mark risk-free rate for the time value of money.

 

[249]   In this approach the Becker Report assessed the risk of Obligors defaulting at the three to four year historic McKesson Canada write-off experience of 0.0440% of sales. Similarly, in this approach the Becker Report assessed the prompt payment dilution risk at the three to four year historic McKesson Canada prompt payment discount experience of 0.5324% of sales.

 

[250]   Adding these three factors together (after first adjusting the 30 day CDOR Rate for the McKesson Canada DSO over the same period) the Becker Report arrives at an arm’s length discount of 0.8073%.

 

Comparable Transaction Approach:

 

[251]   The second valuation approach in the Becker Report is a comparable transaction approach. For this purpose only one arm’s length transaction is considered, that involving McKesson Canada and TD Factors. Dr. Becker considered that an actual arm’s length transaction involving McKesson Canada factoring its receivables to an arm’s length party had a much greater degree of reliability as a starting point than third party agreements or other McKesson Canada financing agreements could have since it was so much closer to the RSA in nature and risk.

 

[252]   The Becker Report calculates the net discount rate in McKesson Canada’s TD Factors transaction as 0.3376%. Since the TD Factors agreement did not specify that TD Factors took any prompt payment discount dilution risk, an upwards adjustment equal to McKesson Canada’s three to four year historic prompt payment discount experience of 0.5324% was made. The Becker Report’s comparable transaction approach concludes that an appropriate arm’s length Discount Rate for the RSA using the TD Factors transaction as the comparable benchmark starting point discount rate would be 0.8700%.

 

[253]   The Becker Report places the appropriate range for an arm’s length discount rate for the RSA between 0.81% and 0.87% (being the discount rates of each of its two approaches), and identifies the midpoint as 0.8386% which is then used as the single discount  rate. The report observes that, had this rate been used in the RSA, McKesson Canada’s pre-tax profit margins would have remained within their historic ranges.

 

[254]   I am not at all satisfied that the TD Factors transaction was sufficiently comparable to the RSA to serve as an appropriate starting point for the Becker Report’s comparable transaction approach, much less to be the only comparable transaction even to be considered. While it was a transaction reflective of McKesson Canada’s creditworthiness and it did involve a transfer of some of McKesson Canada’s receivables, it was for a much shorter term, implemented for a very specific and different purpose, and did not involve the purchase by TD Factors of McKesson Canada’s entire receivables pool. For these reasons, I do not rely in any way upon the Becker Report’s comparable transaction approach.

 

  (g)  The Finard Expert Report

 

                   Structured Finance Approach:

 

[255]   The Finard Report’s structured finance analytical approach to estimating an arm’s length Discount Rate for the RSA was based upon a review of available public bond market rating and credit default loss data. Moody’s back-tests the accuracy of their ratings over a period exceeding 20 years. These show a high degree of correlation between credit ratings and credit default loss rates over periods ranging from one to five years.

 

[256]   The Finard Report looked at one year numbers even though the RSA has a five year term. This recognized that there are several risk mitigation provisions in the RSA that would permit an earlier termination of the RSA. Also, the Moody’s data shows risk of default increases significantly between year one and year five; hence this approach is more conservative.

 

[257]   The McKesson Canada receivables pool’s five year average write-offs to sales was 0.043%. Conservatively using the five-year McKesson Canada write-offs to sales figures, the Finard Report identified that this number, 0.043%, fell between Moody’s one year credit loss percentage by rating for A rated corporate bond issuers and Baa rated corporate bond issuers (or more than five credit ratings above Moody’s “speculative grade” rating which had a credit default history about 90 times higher).[61]

 

[258]   The Finard Report then uses TDSI’s data on credit risk spread by rating and notes that the credit risk spread for an A rated issuer in December 2002 was 0.50%, and for a Baa rated issuer was 1.00%. The Report comes up with a weighted average of 0.68%.

 

Attribute Analysis Approach:

 

[259]   In the Finard Report’s Attribute Analysis approach to estimating an arm’s length Discount Rate for the RSA, a dynamic rolling average approach to the actual and ongoing performance of the receivables was taken to be normative market practice in arm’s length financing transactions that are similarly asset based. The Finard report’s opinion is that the terms and conditions of an arm’s length transaction would have a variable approach to computing DSO and to address prompt payment discounts. Each should be computed based on a four settlement period rolling basis.

 

[260]   In computing a servicing discount, the Finard Report’s Attribute Analysis approach used the $9,600,000 annual fee paid by MIH under the Servicing Agreement and calculated a servicing discount each settlement period based upon the average amount of receivables computed dynamically on a four settlement period basis.

 

[261]   The Finard Attribute Analysis rejected the idea of any discount attributable to MIH’s funding costs (the so-called interest discount) or to the accrued rebate dilutions risk.

 

[262]   The Finard Attribute Analysis approached the issue of loss discount credit risk by (i) taking a dynamic approach to the Loss Discount based upon a rolling 12 month average of write-offs to sales of McKesson Canada’s receivables, and (ii) by introducing a general reserve of 21% of the purchase price of the receivables to deal with the risk of incremental risk beyond the loss discount in the RSA.

 

[263]   The use of a 12 month average of write-offs to sales for the loss discount was chosen because it is consistent with the RSA providing that the loss discount is to be recalculated each January 1^st.

 

[264]   I am troubled by the introduction of a general reserve as enhanced credit protection to MIH on purchased receivables which MIH does not enjoy directly, or even indirectly, in the RSA. There may be significant reserves in securitizations, in secured loans, and in other asset backed loans, but the RSA gives no such security or protection to MIH once it has purchased any particular receivables. I therefore do not see how such a reserve could be appropriate from an arm’s length comparability point of view. I am also somewhat troubled as to whether this can be done in support of an adjustment under paragraphs 247(2)(a) and (c). Given these doubts, which I do not have to resolve in this case given the rest of my reasons, I will not be relying upon the results of the Finard Report’s Attribute Analysis approach.

 

  (h)  The Glucksman Expert Report

 

[265]   In addition to critiquing the Discount Rate approaches in the TDSI Report (and in the PwC Report), the Glucksman Report computes an Affirmative Estimate of an arm’s length Discount Rate for the RSA.

 

[266]   The Glucksman Affirmative Estimate approach dealt with credit risk by introducing an 18.5% reserve. The report expressly regarded the RSA as best analyzed relative to accounts receivables securitizations. It selected 18.5% as the appropriate reserve based upon that having been used in a US$950,000,000 McKesson U.S. receivables securitization.

 

[267]   For the same reasons as described above with respect to the Finard Attribute Analysis’ introduction of a reserve that is not directly or indirectly reflected in the RSA, I will not be relying upon the results of the Glucksman report’s Affirmative Estimate approach to estimating an arm’s length Discount Rate for the RSA. Similarly, the Glucksman Report’s Affirmative Estimate approach is viewed entirely through the lens of accounts receivable securitizations and, for that reason, I will not be relying upon any portion of the Glucksman Report’s Affirmative Estimate approach. For these reasons, I will dispense with summarizing the rest of the report’s Affirmative Estimate approach.

 

  (i)  The Other Expert Reports

 

[268]   Further expert reports were filed on consent, without the Court hearing testimony from their authors. As with the PwC Report, little weight can be given to the contents of expert reports written by persons who did not testify in the proceedings. To some small extent they might provide corroboration for the approaches or data of others.

 

9.       The Appropriate Methodology

 

[269]   As is often the case where there is more than one expert and expert report, the Court does not accept the conclusions of any of those experts or their reports in their entirety. However, while some approaches and some information, estimates or components of their analysis may be unpersuasive or rejected, overall the Court is informed by all of their testimony and the information they provided relating to the opinions that they arrived at and, as can be seen, has relied upon parts of their opinions and some of the factual information they relied upon.[62]

 

[270]   The Court is of the view that the most appropriate and proper approach in this particular case is to follow the structure of the RSA that the McKesson Group chose to enter into and to approach the pricing issues largely as TDSI (and PwC) did, and consider whether the terms and conditions which affect the Discount Rate pricing differ from what arm’s length terms and conditions would be expected to provide, in order to adjust the amounts best as I can to reflect the helpful evidence on these issues, including the expert opinion evidence, before me.[63]

 

10.     Analysis of Transfer Pricing Issue

 

[271]   This was not a securitization transaction. A securitization is generally an off-balance sheet debt financing, via a thinly capitalized special purpose entity, that accesses low rate investment grade financing via a structured finance product that incorporates risk minimization features including support from the seller of the existing cash flow stream. The Appellant’s evidence is 100% consistent that this was not a securitization transaction, nor was there any apparent financial or business reason for McKesson Canada to be interested in a securitization transaction. Any comparison of any aspect of this transaction with a securitization needs to be cautiously approached with this in mind.

 

[272]   The RSA was a five year facility. This was not five year money any more than it was simply 30 day money. Any suggestion that any aspect of it could be equated to comparable terms of five year debt, medium term debt, or other long term debt needs to be cautiously approached.

 

[273]   Similarly, the RSA had a maximum receivables pool at any time of $900,000,000. Any suggestion when making comparisons to other transactions that this represents that the RSA reflected $ 900,000,000 of exposure, unless and until MIH ever hit that maximum would similarly need to be very cautiously approached and thought through.

 

[274]   I find as a fact that the predominant purpose and intention of McKesson Canada participating in the RSA and related transactions with the other McKesson Group members was not to access capital or to lay off credit risk. Those were results of the transactions but did not motivate them. The purpose was to reduce McKesson Canada’s Canadian tax liability (and therefore McKesson Group’s worldwide tax liability) by paying the maximum discount under the RSA that McKesson Group believed it could reasonably justify. For the McKesson Group this appears to have been much more of a tax avoidance plan than a structured finance product. No reason was ever given for wanting to transfer risk to Luxembourg.

 

[275]   There is certainly nothing wrong with taxpayers doing tax-oriented transactions, tax planning, and making decisions based entirely upon tax consequences (subject only to GAAR which is not relevant to this appeal). The Supreme Court of Canada reminds us regularly that the Duke of Westminster is alive and well and living in Canada. However, the primary reasons and predominant purposes of non-arm’s length transactions, whatever they may be in any given case, form a relevant part of the factual context being considered. For example, if neither side has a business purpose or need to do a particular non-arm’s length transaction, it will probably not be particularly persuasive to try to argue that particular terms, conditions, provisions, or approach reflect the particular business need of either party.

 

[276]   The maximum amount deductible in Canada by McKesson Canada is limited to what an arm’s length person would agree to pay for the rights and benefits obtained. The Appellant says it did not exceed that limit. The Respondent says they exceeded it by more than 100%. This is the only question that the Court is called to decide.

(a) The Discount Rate

 

[277]   As already described, the RSA provides that the Discount Rate for each purchase of receivables is the sum of (i) the Yield Rate on the first business day of the relevant settlement period, (ii) the Loss Discount (iii) the Discount Spread.

 

(i)      The Yield Rate

 

[278]   The Yield Rate is the only fully floating component of the Discount Rate in the RSA. In contrast, both the Loss Discount and the Discount Spread are fixed. The Loss Discount is recalculated annually or earlier at MIH’s request.

 

[279]   There is no dispute on the evidence that the 30 day CDOR rate is the appropriate base line risk-free rate. I accept that, and I accept that using that rate as of the first business day of each 28 day settlement period (ignoring for the moment that the RSA was signed on December 16, 2002, nine days after the end of McKesson Canada’s Accounting Period 9 of 2003, on December 7, 2002, leaving only a 19 day initial settlement period)  is well within the range of what two arm’s length parties, entirely adverse in interest on pricing and risk-related terms and conditions, would agree to as both acceptable and reasonable.

 

[280]   The CDOR rate as of December 16, 2002 was 2.79% per annum. There was no evidence that this changed materially in the remaining two or three relevant Accounting Periods in McKesson Canada’s 2003 year. For purposes of this appeal of McKesson Canada’s 2003 year, I will assume the CDOR rate did not change materially.

 

[281]   The CDOR rate is expressed as an annual rate and this needs to be adjusted to reflect that the receivables can be expected to be collected over a much shorter period than a 365 day year. I accept that using the accounting concept of Days Sales Outstanding or DSO for the collection period of McKesson Canada receivables is an appropriate proxy or measure for this purpose.

 

[282]   However, given my observations above on the calculation of DSO by McKesson Canada and TDSI for the RSA, I do not accept that using a fixed DSO of 31.73 days throughout the term without regard to changes to McKesson Canada’s actual DSO at the relevant time (whether resulting from changes in McKesson Canada’s sales or changes in its customers’ payment patterns) is what two arm’s length parties, adverse in interest as to pricing and risk-related terms and conditions, would agree to.

 

[283]   I find that arm’s length parties would choose to incorporate a floating approach to DSO averaged over some period, say three to four months or Accounting Periods[64] and would not accept the risk of fixing the DSO for the entire term of the RSA. Notably the very concept of DSO is measured and tracked over successive periods of time. This is also supported by the evidence of experts Becker and Finard. Given that the RSA uses a four month rolling average in its definitions of Loss Ratio and Delinquency Ratio, I find a rolling four Accounting Period average appropriate in this appeal. In any event, I do not have any helpful evidence to support a fixed approach as the Appellant has not adequately explained how the risk of change in the DSO was factored into the Discount Rate pricing,[65] why the risk of change in the DSO is not relevant, nor provided other supporting evidence of how this Court can price or adjust for the risk of DSO change.

 

[284]   TDSI calculated McKesson Canada’s DSOs for each Accounting Period of McKesson Canada’s 2001, 2002 and 2003 fiscal years as part of its engagement. The average DSO per period over this term prior to the RSA as computed by TDSI was 32.00 days.

 

[285]   While this is stated accurately in the text of the TDSI Report, the TDSI backup schedules state that the average was 31.73 days. This is not correct; if one works the average out from the schedule of numbers it is 32.00 days. The fact that the average shown in the schedule is 31.73 days, the same number TDSI uses after its rough adjustment for the significantly shorter DSO on the maturing receivables pool of $460,000,000 initially purchased, certainly raises questions.

 

[286]   TDSI then recognized, in helping the Appellant fix its DSO for purposes of the RSA computations, that the DSO of 32.00 days would not be appropriate given that the initial $460,000,000 purchase of the existing receivables pool on December 16, 2002 was a mature pool of receivables. TDSI estimated that a mature pool in the circumstances should be expected to be paid within one-half of the existing DSO, or within sixteen days. That seems to be a sensible and reasonable estimation.

 

[287]   However, rather than use a 16 day DSO for the initial receivables purchase’s Discount Rate, TDSI instead averaged these “missing” sixteen days across the five-year term and simply reduced the 32 day DSO to 31.73. I do not accept that is appropriate nor that it is an arm’s length approach to this issue. This allowed MIH to underpay McKesson Canada at the outset by virtue of a significantly overstated Discount Rate, and it would only accrue back to McKesson Canada over the five years and without interest. I am not persuaded that arm’s length parties to a financial transaction would agree to deal with the issue on that basis alone. As has been noted, this also created a corresponding five year Canadian income tax timing benefit to McKesson Group.

 

[288]   I conclude that arm’s length parties to a financial transaction would accurately account for the difference between the initial purchase’s estimated 16 day DSO and TDSI’s 32 day DSO for new receivables. This will materially impact each of the components of the Discount Rate that are annual rates that need to be DSO adjusted. These are the Yield Rate and components of each of the Loss Discount and the Discount Spread. This would not be an insignificant transfer pricing adjustment for McKesson Canada’s 2003 year under appeal.

 

[289]   I do not, however, accept TDSI’s computation of a 32 day DSO or its backup schedules to its report. I prefer the McKesson Canada DSO schedules which show that the DSO as computed by the company in its ordinary course averaged less than thirty for the two financial years (2002 and 2003) preceding the RSA, and in only one Accounting Period of the 18 preceding the RSA exceeded 30 (being 30.5). This is consistent with Mr. Brennan’s evidence that the DSO was and remained in the 30 day range. I conclude that arm’s length parties would use a four month rolling average DSO and that the best evidence of this was 30.0 days throughout.[66]

 

[290]   Adjusting the 30 day CDOR rate of 2.79% per annum for a 30 day DSI results in a Yield Rate of 0.2293 (2.79% x 30 ÷ 365).

 

[291]   The corrected or adjusted Yield Rate for the initial purchase would be half of this. The Yield Rate for the 2003 year in issue, which has three settlement dates and full DSO cycles, will be only 0.1911 (the sum of 0.2293 plus 0.2293 plus 0.1147, all divided by 3). This would spread the “missing” 15 days over the three month period following the initial purchase, and all within the same tax year being the year under appeal.

 

(ii)     The Loss Discount[67]

 

[292]   The Loss Discount in the RSA is intended to account for the risk that the Obligors do not fully pay their receivables. As outlined above, McKesson Canada’s multi-year receivables performance numbers resulted in an average collection of 99.96% of its receivables. This represents 0.04% write–offs to sales.

 

[293]    As described above, the Loss Discount was a fixed 0.23% throughout the year in question and to December 31, 2003.

 

[294]   The Loss Discount was to be recalculated in accordance with the RSA each year thereafter or whenever MIH felt the Designated Obligor receivables ratio had materially changed since last calculated. Effectively, only the Loss Discount attributable to Designated Obligors could be changed under the terms and conditions of the RSA as the Loss Discount attributable to other Obligors was fixed at 0.2380% for the full five-year term.

 

[295]   The evidence confirmed that the fixed 0.23% Loss Discount for 2003 was calculated by the Appellant and TDSI on the same basis as the annual recalculation provided for in the RSA.

 

[296]   The Court does not accept that the terms and conditions of the RSA relating to the Loss Discount of Designated Obligors, or the approach of the McKesson Group and TDSI to computing the Loss Discount of other Obligors, reflect arm’s length terms and conditions. The Appellant has not satisfied the Court on the evidence presented that these terms and conditions of the RSA reflect what arm’s length parties, adverse in interest as to pricing and risk, would agree to in similar circumstances.

[297]   Firstly, the evidence does not permit or support any conclusion being reached on the extent to which a rated company’s bond rating is a reasonably accurate market approach to assessing risk of default on the company’s trade payables for inventory owing to a key dominant supplier. There was little if any evidence to support a sufficiently direct or comparable correlation, though I do accept that one might reasonably expect a degree of relevance. Some evidence I heard confirmed the contrary. The Appellant did not tender any supporting evidence for the credit analysis, ratings or scores assigned by McKesson Canada’s Credit and Collections department of its rated Obligors to support or explain its position. In the circumstances of this trial, I can only conclude that was intentional.

 

[298]   Secondly, I can not reasonably conclude that a company that does not have a bond rating can be assumed to be hiding a bad implicit rating from the public. Many private companies, large or small, do not obtain public bond ratings simply because they have no need or desire to raise money in the bond markets as conventional lenders and sources of funds work fine for them. There is no rational basis supported in the evidence or in common sense that all unrated companies are equivalent credit risks to non-investment grade bond issuers.

 

[299]   The Court’s concerns are further confirmed from a common sense point of view by the fact that the approach taken by McKesson Group and TDSI assigns a going forward credit risk to the receivables of McKesson Canada’s customers from time to time that is many, many, many times higher than the multi-year historic performance of these receivables.

 

[300]   The Court does not accept that the Loss Discount attributable to Obligors other than Designated Obligors would be fixed by arm’s length parties in the manner it was, nor fixed at that same number for a five-year term.

 

[301]   The Court does not accept that an arm’s length seller, would agree to terms and conditions that resulted in a Loss Discount that was almost 600% of its historic receivables write-offs without any significant projected, planned or reasonably anticipated material risk of deterioration of its business, its customers, its receivables, or the Canadian or world economies generally, of which there was no evidence.

 

[302]   The Court does not accept that arm’s length parties would agree that the Loss Discount of Designated Obligors could be recalculated at any time if the buyer thought that the mix of Designated Obligors to other Obligors had changed, but not if the seller thought the mix had changed in its favour, absent some off-setting concession or quid pro quo, of which there was none.

 

[303]   The Court does not accept that arm’s length parties would agree that the RSA’s Loss Discount terms, nor their underlying computations by McKesson Group and TDSI, would not directly take into account at all neither the historical or actual ongoing performance of the receivables pool.

 

[304]   For all of these reasons, I can not accept that the RSA terms and conditions regarding Loss Discount, and the computations of the quantum of the Loss Discount by McKesson Group and TDSI, were arm’s length terms and conditions or resulted in the appropriate arm’s length amounts. It is therefore the role of this Court to estimate what the appropriate amount or range of the Loss Discount should be for purposes of computing the Discount Rate under the RSA. In doing so, this Court is limited to the evidence available to it. This Court can only use an estimation method that is able to be used with the available evidence to arrive at a number or range of numbers. The parties’ choice of evidence may be a constraining factor which may well preclude an arguably more effective or appropriate method for the Court to estimate an arm’s length amount.

 

[305]   There is no magic about past historical data. We can not consider future data, only make reasonable predictions. After the fact we may also now have to assess whether a past assumption about the future, had it been made, would have been reasonable. In doing so, we may be cognizant of what would then have been future data but which is now equally historical. All must be approached carefully by a court. None are determinative, but none are entirely irrelevant considerations.

 

[306]   The RSA was signed at a time when the receivables pool’s write-offs to sales performance had been in the range of 0.04%. This was well known and tracked by McKesson Canada and McKesson Group. The RSA gave MIH an immediately exercisable termination right in the event the pool’s Delinquency Ratio or Loss Ratio increased by specific measures.

 

[307]   Ms. Hooper’s evidence was that these two termination event triggers in particular were designed to effectively stop the transfer of additional receivables once the portfolio does not perform as well as it did in the past. The Delinquency Ratio was designed as an early warning system. Given that delinquencies can be expected to increase in advance of seeing losses increase, the termination right was designed to occur early enough that one is not going to have very material losses. According to Ms. Hooper’s testimony, the Loss Ratio and Delinquency Ratio combined should allow the purchaser to stop acquiring additional receivables in time to not suffer materially higher losses than expected based on past performance. Ms. Hooper testified that she and her team at TDSI looked at both the historical loss and delinquencies in the McKesson Canada receivables pool as part of its engagement in preparing the TDSI Report.

 

[308]   I do not necessarily accept the TDSI Report’s opinions on the reasonableness, normalcy or arm’s length nature of these two termination triggers in the RSA. Indeed, I would expect they might suffer from the same shortcomings as affects the rest of the TDSI Report, which is primarily that the RSA is not a securitization and is in that respect outside the expertise of Ms. Hooper and her group. In any event, given that these two ratios are defined in the RSA to include McKesson Canada financial information that is not in evidence, or at least certainly not adequately explained in the evidence, and that these defined ratios and their volatility leading up to the RSA were not put in evidence, I can not reach the conclusion that I am satisfied with the TDSI’s Report’s conclusions on their terms.

 

[309]   However, I fully accept Ms. Hooper’s explanation of their purpose and effectiveness as designed. That is, I find that the purpose and effect of the Delinquency Ratio termination event trigger and the Loss Ratio termination event trigger in the RSA were designed and fully expected to limit MIH’s risk of purchasing any day’s receivables from McKesson Canada that could be expected to have materially higher losses than had been experienced on the pool historically.[68]

 

[310]   The historic loss performance on McKesson Canada’s receivables pool was in the range of 0.04%. I conclude from all of this that a notional arm’s length MIH would have been able to and would have terminated its obligations under the RSA before it was obligated to purchase receivables that would have a materially higher credit loss risk than something in the range of 0.04%.

 

[311]   Allowing for a 50% to 100% increase as an extremely generous interpretation of what Ms. Hooper could have meant by material (in part to compensate for the lack of elegance in this approach), I find that a notional arm’s length MIH’s credit loss risk on its continuing purchase of receivables is that, at some future point in the RSA term (but not in the very short term), it could have purchased about four months of receivables with an anticipated write-offs to sales number in the range of 0.06% to 0.08%. These lesser quality receivables would only be expected to have been purchased in the last four months of the RSA prior to termination. Those bought on December 16, 2002 and for the other months prior to the four months preceding termination could continue to be expected to be of a better quality.

 

[312]   Using this approach, the Court concludes that a Loss Discount component of the Discount Rate in the range of 0.06% to 0.08% is at the generous end of what a notional arm’s length MIH and McKesson Canada would agree to.

 

[313]   This range is consistent with the number arrived at by Mr. Finard’s structured finance approach. That approach identified that the 0.04% historic write –offs to sales number for McKesson Canada’s receivables pool was comparable to Moody’s published information for companies rated between A and Baa, which in turn had credit risk spreads according to TDSI of 0.50% and 1.00% per annum, and was computed on a weighted average basis by Mr. Finard at 0.68%. Once adjusted for a DSO of 30 days, a 0.68% annual credit spread reflects a discount of 0.06%.

 

[314]   For these reasons, the Court finds based upon what evidence was provided that an arm’s length Loss Discount for purposes of the RSA would be in the range of 0.06% to 0.08%.[69]

 

(iii)    The Discount Spread

 

[315]   The discount spread in the RSA was calculated by McKesson Group and TDSI as the sum of four different components.

 

1.       Servicing Discount

 

[316]   The Court does not accept the TDSI Report’s position regarding the discount needed to reflect the potential cost of choosing to engage a new servicer to replace McKesson Canada as servicer of the already purchased receivables.

 

[317]   The Servicing Agreement sets out a fee payable of $800,000 monthly to the servicer and appoints McKesson Canada as the initial servicer. By its terms, the Servicing Agreement applies to a replacement servicer, assuming a replacement servicer would agree to it.

 

[318]   Mr. Reifsnyder was of the opinion that the $9,600,000 annual fee provided in the Servicing Agreement was substantially rich enough that it could possibly fully cover servicing costs even if a replacement servicer needed to be brought in for a short period of time at that level. Mr. Reifsnyder had been involved in transactions where servicers needed to be replaced.

 

[319]   I do not accept TDSI’s unexplained 25% chance that a replacement servicer would need to be appointed. That is entirely unsubstantiated. TDSI begins from a less than 10% prospect of a rating downgrade termination event.

 

[320]   Even upon the occurrence of a termination event followed by termination, I do not accept that it is a given that a notional arm’s length MIH would invariably exercise its right to appoint a replacement servicer. The evidence is that McKesson Canada had very good and successful collection policies, practices and results. A number of the termination events listed in the RSA could occur without any related impact upon McKesson Canada’s continuing servicing abilities.

 

[321]   The appointment of a replacement servicer for a short period of time, a period of between 30 days (the DSO) and 90 days, being how long TDSI thought would be needed to wind up the portfolio once no new receivables were being purchased, would have to be weighed against disruption in the customer relationships which may further delay and hinder payment, learning time and inefficiencies for a new servicer to get up to speed, and cost considerations (hard and soft costs, direct and indirect). The prospect of having to appoint a replacement servicer would have been assessed by a notional arm’s length MIH at much less than the 25% assumed by TDSI or the 40% assumed by PwC. Based upon all of the evidence above, I would estimate that the likelihood would have been significantly less than 10%.

 

[322]   I accept Mr. Reifsnyder’s opinion based upon his experience with replacement servicers and servicing in securitizations, that it was quite possible the servicing fee set out in the Servicing Agreement was sufficient to fund a replacement servicer if one was needed.

 

[323]   McKesson Group and TDSI calculated the servicing discount using a replacement servicer cost of 2% of the face amount of receivables. The only support for this was the range of 1% to 3% of face obtained by TDSI from one third party provider, perhaps a major accounting firm. The other evidence on the cost of replacement services according to the TDSI Report, is that Bell Canada recorded 1% and Telus 2% on their Canadian receivables securitizations. PwC, a major accounting firm, used replacement servicer fees in the range of 0.8% to 1.2%. Moody’s and Standard & Poor’s both put replacement servicers in the 1% range of the face amount of receivables.[70]

 

[324]   I find that a reasonable maximum replacement servicer fee would be 1% of the receivables to be collected in the case of a notionally arm’s length RSA.

 

[325]   MIH was paying $9,600,000 annually to the servicer under the Servicing Agreement which works out to $800,000 per month. Given that the DSO of the purchased receivables was also in the one month range, this fee represents about 0.17% of the $460,000,000 RSA utilization in 2003 by McKesson Canada. This fee under the Servicing Agreement has not been challenged or reassessed and is not directly in issue in this appeal. On the basis of the fees payable under the McKesson Group’s Servicing Agreement, which were not challenged by CRA, and which were supported by TDSI in its supplemental report, I accept that a notional arm’s length RSA and Servicing Agreement would allow a basic servicing discount to be financed out of MIH’s discount under the RSA equal to 0.17%.[71]

 

[326]   If a new replacement servicer was appointed, this fee would no longer be paid to McKesson Canada and would be paid to the new third party servicer, in accordance with the Servicing Agreement.

 

[327]   According to Mr. Reifsnyder, this may require no additional outlay beyond the Servicing Agreement amounts. Following the McKesson Group and TDSI approach, a replacement servicer, if appointed, would only be paid once to collect and wind down the receivables pool in orderly fashion following termination of the RSA by MIH. Assuming the receivables pool might have reasonably been expected to have increased to $500,000,000, a 1% replacement servicer fee payable by the Purchaser would be $5,000,000. In my opinion, that is the absolute, outside maximum total dollar amount of discount that a notionally arm’s length McKesson Canada would be agreeable to pay a purchaser for the replacement servicer risk.[72]

 

[328]   After accounting for the $800,000 already available each month, MIH would need to pay no more than an additional $4,200,000 upon termination. If this additional $4,200,000 were to be simply collected over the five-year term, this would only require less than a .02% discount assuming constant RSA utilization at $460,000,000. This would not account for MIH’s risk that a replacement servicer might need to be appointed before the last period of year five. On the other hand, it also would not account for the fact that there is much less than a 100% chance of a replacement servicer being needed.

 

[329]   If the $4,200,000 is to be fully collected in the first 12 months, the additional discount needed for replacement servicer risk under the McKesson Group/TDSI approach is 0.08%. If this was to be fully collected over the first half of the RSA term, being 30 months, the additional discount needed would be .04%.

 

[330]   Even if a notional arm’s length MIH wanted to recover this $4,200,000 all in the first three settlement periods of the RSA, being those occurring in McKesson Canada’s 2003 taxation year, this means it would need to enjoy an additional discount of $1,400,000 in each period. Assuming the projected RSA utilization remained at $460,000,000 for those three months, this would account for a replacement servicer cost discount of 0.3% for those three months only.

 

[331]   Based upon the numbers above, I would estimate that it is reasonable to conclude that a notional arm’s length McKesson Canada and a notional arm’s length MIH would agree to allow MIH to fully fund and cover the potential additional cost of a replacement servicer, should one need to be appointed, over a period ranging no shorter than from the first 12 months of the RSA (which would require an additional discount of .08%) and to no longer than the first 30 months of the RSA (which would require an additional discount of .04%).

[332]   This approach to using the RSA discount to fund the servicing fee under the Servicing Agreement, and to fund the potential replacement servicer risk, does not constitute anything more than changing the quantum of a term or condition of the parties’ RSA given that the Discount Spread only consists of a numerical amount. It is certainly well within the range of what is permitted under paragraphs 247(2)(a) and (c).

 

[333]   Adding the servicing fee discount based upon the Servicing Agreement of 0.17% which Mr. Reifnsyder believes might be sufficient to also cover off any replacement servicer risk, to the maximum of the McKesson Group/TDSI approach’s range of replacement servicer fee discounts of .04 to .08%, the Court estimates, based upon what evidence was tendered, that the appropriate range of servicing discount in the Discount Spread of a notional arm’s length RSA would be in the range of 0.17% to 0.25%.

 

2.       Prompt Payment Dilutions Discount

 

[334]   It is not clear why the parties to the RSA provided that prompt payment discounts are not treated as deemed receipts but are instead at the purchaser’s risk. Regardless, that is what the RSA provides and I accept that arm’s length parties might agree to such a term in a similar receivables financing transaction. The issue is therefore only whether the prompt payment dilutions risk has been accounted for in the Discount Spread component of the Discount Rate on arm’s length terms.

 

[335]   The historic prompt payment discount levels are very consistent according to the evidence at 0.5% or 0.53% of sales. TDSI tested this on an annual basis over several years and came back with 0.5%. The Becker Report independently arrived at a three to four year prompt payment discount experience of 0.5324% of sales.

 

[336]   I do not accept that arm’s length parties would agree to a fixed discount spread for the five-year term to address in a balanced fashion the risk of change to prompt payment discounts taken by McKesson Canada’s customers. Specifically, I do not accept that arm’s length parties adverse in interests as to risk and pricing would agree to the 20% TDSI buffer or to the PwC 5% cushion.

 

[337]   If arm’s length parties were to agree to a transfer of prompt payment discount risk (upside and downside) to the purchaser, I conclude that for the initial purchase in December 2002 this would be based upon the historic 0.5% to 0.53%of sales performance to date. I do not accept that an arm’s length seller would walk away from the possibility of favourable variance if it was agreeing to give the buyer a 5% or 20% cushion. I do not accept that it would be fixed for the five-year term. I do not accept that either party would not require the other party to also factor the effect of more prompt payments upon DSOs for all purposes of the Discount Rate.

 

[338]   In my estimation, based upon the evidence, arm’s length parties would instead agree to virtually remove the risk of change during the term of the RSA in the levels of prompt payment discounts by adopting a three or four month, or annual floating dynamic prompt payment component to the Discount Spread component of the Discount Rate to fully capture and reflect the risk of change. This approach is supported by the evidence of Dr. Becker and Mr. Finard. The Reifsnyder Report was not concerned with risk of change to prompt payment discount levels because trade receivables were not considered variable as to prompt payment discount participation. This would again be permitted by paragraphs 247(2)(a) and (c).

 

[339]   Using this approach, the Discount Spread component attributable to prompt payment dilutions would be 0.5% to 0.53% for the initial December 2002 receivables purchase. Since this number had proven to be very consistent, and there was no evidence to suggest that was expected to change, or had changed, in any material way in the following few months, I further estimate that arm’s length parties would have agreed to use a number in the range of 0.5 to 0.53% to reflect that prompt payment discounts were borne by the purchaser under the RSA for the initial purchase in December 2002 and for the remainder of McKesson Canada’s 2003 year.

 

3.       Accrued Rebate Dilutions Discount

 

[340]   I do not accept that the accrued rebate dilutions risk warrants any material discount. The Appellant’s expert reports did not support one. This thought was that of McKesson Group and TDSI. Their approach was followed in the PwC Report which went on to describe the accrued rebate risk as an expected loss that MIH would suffer. Glucksman did consider it only reluctantly preferring instead to substitute a reserve or pledge type approach. I have noted my thoughts on the shortcomings of TDSI’s approach to this above.

 

[341]   I certainly can not imagine that McKesson Canada would agree with a notional arm’s length purchaser (absent a significant corresponding concession or quid pro quo of which there was none in evidence) to a discount that either (i) reflected a full recovery for the notional arm’s length purchaser based upon an assumption that all of McKesson Canada’s customers would exercise a claim to set off their accrued rebate entitlements, and this at a time when the total accrued rebates were at an historic high, or (ii) that assigned a credit risk spread to McKesson Canada equivalent to junk bond issuer status.

 

[342]   There was no evidence that McKesson Canada’s customers had ever claimed, threatened or even asked to set off their accrued rebates against their outstanding payables to McKesson Canada.

 

[343]   In the circumstances described herein and the evidence before the Court, it is my estimation that McKesson Canada would not agree with a notional arm’s length purchaser to any material discount to reflect the mere possibility of a rebate set-off claim being made and for which McKesson Canada would not in fact indemnify the receivables purchaser. There would thus be no need to assign a McKesson Canada credit risk spread to such an eventuality which is unsupported on the evidence.

 

[344]   This is supported by the express evidence of Messrs. Becker and Finard and is consistent with accrued rebate risk not being identified, much less quantified, as a cash flow dilution risk of the RSA in the Reifsnyder Report.

 

[345]   This is also further supported to an extent by the fact that McKesson Canada was not required to segregate collections, and was permitted to commingle collections under the RSA (absent termination). That is, no credit risk was recognized in respect of McKesson Canada’s ability to pay amounts owing, even very significant amounts. The McKesson Group did not consider there to be a material risk of McKesson Canada insolvency or bankruptcy, nor any other financial risk arising from the commingling of funds.

 

4.       Interest Discount

 

[346]   The interest discount used by McKesson Group in the RSA was intended to provide MIH with a return from the discounted purchase of receivables, in addition to all of the above amounts, equal to an assumed cost of funds (that it did not in fact bear) equal to the cost to a below investment grade borrower that borrowed 100% of the receivables purchase price by issuing its junk bonds in the market. I find this completely unacceptable, unreasonable, unsupported on the evidence, and a term that would not be agreed to by McKessson Canada in a similar receivables financing transaction with an arm’s length party adverse in interest as to risk and pricing.

 

[347]   As a general rule, the value of an asset to be sold is not generally affected by a particular purchaser’s cost of funds. Generally, a business or an investor with cash or a low cost of funds can profitably make less risky investments with a lower nominal return on investment than can a person with a high cost of funds. A purchaser’s cost of funds does not decrease the value of the asset it wishes to buy or the investment it is considering. Rather, it simply determines whether that particular purchaser can make the purchase or investment profitably, and if so, how profitably.

 

[348]   There was no satisfactory evidence tendered that would suggest that McKesson Canada was driven to seek receivables financing from a high cost of funds/high cost factoring company and not a better funded/lower yield/lower cost major financial player described in the taxpayer’s own evidence. I was not, however, provided with evidence of the cost of capital associated with receivables factoring by major well-funded players.

 

[349]   In the circumstances of the McKesson Group RSA in this case, and based upon what evidence I do and do not have, I find that McKesson Canada would not agree to sell its receivables at a Discount Rate that incorporated an interest discount to reflect its notional arm’s length purchaser’s cost of funds at the level set by McKesson Group in the discount spread component of the RSA’s discount rate of 0.4564%. Based upon the evidence in this case, I estimate that, for the year in issue, in a notional arm’s length transaction, a notional arm’s length McKesson Canada would only agree to an interest discount of between 0.0% and 0.08%, which latter number reflects a 30 day DSO adjusted rating-derived credit risk spread for a company having the same rating as McKesson U.S., according to TDSI.

 

(b) Summary of Court’s Estimate of Discount Ranges

 

[350]   Tabulating the above Discount Rate components arrives at the following:

 

Yield Rate:                                                  0.2293%

Loss Discount:                                            0.06 – 0.08%

Servicing Discount:                                    0.17 – 0.25%

Prompt Payment Dilutions Discount:                  0.5 – 0.53%

Accrued Rebate Dilutions Discount:           0

Interest Discount:                                        0 – 0.08%

          _______________________________          ________________

          Total Discount Rate Range:                        0.959% to 1.17%

 

[351]   Using the adjusted Yield Rate of 0.1911 for the 2003 year in issue to reflect the much shorter expected repayment period for the initial December 2002 $460,000,000 receivables purchase,[73] this Discount Rate range is reduced to 0.92% to 1.13%. This range will be even further reduced once the 15 or 16 day DSO is also applied to the DSO adjusted components of the Loss Discount and Discount Spread.

 

11.     Conclusion on Transfer Pricing Adjustment

 

[352]   The Court has concluded that its best estimate of the range of Discount Rate to which arm’s length parties to a notional arm’s length RSA would agree is between 0.959% and 1.17%.

 

[353]   The taxpayer has not been able to establish with sufficient credible and reliable evidence, that the RSA Discount Rate of 2.206% was computed based upon arm’s length terms and conditions.

 

[354]   The evidence does not show that the Discount Rate used by the Minister of National Revenue (the “Minister”) in the reassessment of 1.013% was below a Discount Rate computed on arm’s length terms and conditions for a notional arm’s length RSA. The Court can not conclude on all of the evidence that the reassessment was incorrect as it was within the arm’s length range determined by the Court.

 

[355]   In any event it is not necessary to fix a particular point within the determined range as the arm’s length transfer price as, importantly, the taxpayer’s evidence does not rise to the level of making out a prima facie case that “demolishes” the key assumptions of fact made by the Minister that support the reassessments.[74]

[356]   Assumption (v) in paragraph 28 of the Amended Reply is that the Discount Rate that would have been agreed to had the Appellant and MIH been dealing at arm’s length would have been established of a rate no greater than 1.0127%. The taxpayer has not been able to discharge the burden and onus upon it of showing that the amount of the reassessment is incorrect. The shortcomings of the TDSI Report, the Reifsnyder Report, and the supporting testimony regarding both, were obvious and apparent and did not require contrary evidence from the Respondent to make them evident. For this reason, the taxpayer’s appeal with respect to the transfer pricing adjustment is dismissed.

 

[357]   This is an appropriate result. It would not be appropriate for this Court to order the Minister in a case such as this to reconsider and to reassess at the high point of the range of arm’s length Discount Rates (1.17%). That would reward overreaching taxpayers who would then count on the court process to ensure they enjoyed the highest permissible transfer price. This would encourage the poor use of public resources and expenditures. In contrast, in transfer pricing disputes which, as here, often involve very large amounts, the taxpayer’s costs can be less than the value of even a slight variance in the underlying price of the inputted asset or service. Taxpayers would be economically encouraged to use the Court to ensure they get their maximum transfer price by choosing one that is likely to exceed it.

 

[358]   Further, the Discount Rate range with respect to the year in issue is less than 0.959% to 1.17%. Estimating that the midpoint of the 2003 range is the appropriate arm’s length Discount Rate, and after making the further needed DSO adjustments to components of the Loss Discount and the Discount Spread described in paragraph 288 above, it appears that in any event the arm’s length Discount Rate for the 2003 year in issue as determined by the Court is less than the rate used by the Minister in the reassessment.

 

12.     Timeliness of Part XIII Assessment of McKesson Canada

 

a) The Issue

 

[359]   The vicarious assessment of McKesson Canada is for its failure to withhold and remit to CRA on the benefit it paid to its parent (and sole shareholder), MIH, via the transfer of receivables at an overstated Discount Rate which resulted in it having given away some of its assets to its parent/shareholder.

 

[360]   The amount of this benefit is deemed to have been a dividend paid by McKesson Canada to MIH upon which MIH is subject to Canadian Part XIII non-resident withholding tax.[75] Under the Canada-Luxembourg Treaty[76] (the “Treaty”) (which Respondent only admits for purposes for this appeal does apply) the non-resident withholding tax rate payable on dividend income received from a Canadian by a Luxembourg resident is reduced from 25% to 5%.

 

[361]   Distinct from MIH’s liability for Part XIII tax under the Act on its Canadian-sourced dividend income, McKesson Canada as payor is obliged under the Act to withhold from MIH as payee and remit to CRA on behalf of the non-resident, an amount equal to the amount of non-resident withholding tax payable by the non-resident payee. The Canadian payor, McKesson Canada, is itself liable under Part XIII of the Act for an amount equal to the amount that should have been withheld from the non-resident, and remitted to CRA by it but was not. The withholding obligation exists to facilitate enforcement and collection in Canada against Canadian payors without needing to pursue non-resident payees. If this collection mechanism is not complied with, the subsection 215(6) vicarious direct liability of the Canadian payor for an equivalent amount then further serves this same purpose. The Act provides that the Canadian can seek indemnity from the non-resident.

 

[362]   The transfer pricing reassessment of McKesson Canada for additional Part I income tax under the Act was issued by CRA on March 25, 2008.

 

[363]   The Part XIII assessment of McKesson Canada for its vicarious liability for an amount equal to the amount it should have withheld from, and remitted on behalf of, MIH was issued by CRA on April 15, 2008.

 

[364]   CRA did not ever assess non-resident withholding tax against MIH with respect to the non-arm’s length RSA Discount Rate benefit by imposing Part XIII non-resident withholding tax on MIH’s Canadian sourced dividend income.

 

[365]   There was no evidence that there was a material amount of Luxembourg tax payable by MIH on its profits under the RSA. The only evidence is that in its 2003 short year the Canadian tax avoided by McKesson Group was US$4,500,000 and that some form of Luxembourg tax was expected to be payable in the amount of US$29,000. I was similarly not given any evidence that there was a material or any amount of Luxembourg tax payable by MIH on its deemed dividend under the Canadian Act.[77]

 

[366]   The Treaty has an express and specific five calendar year limitation for assessing tax on certain transfer pricing adjustment income in specific circumstances. CRA was mindful of this five-year period in issuing the transfer pricing adjustment reassessment of McKesson Canada, beating that date by a few days. For some reason that was not explained, the Part XIII assessment was issued to McKesson Canada by CRA three weeks later and outside the five-year period, assuming the Treaty applied to such an assessment. The Part XIII issue in this appeal is whether Article 9 of the Treaty applies to the assessment of McKesson Canada by the Canadian tax authorities for its failure to remit to CRA the amount it was required by the Act to withhold from MIH. It does not involve an assessment by CRA of MIH for Canadian tax.

 

b) The Provisions of the Income Tax Act and the Treaty

 

[367]   The relevant portions of the provisions of the Act are as follows:

 

Part 1, subsection 15(1):

 

15(1) Benefit conferred on shareholder -- If, at any time, a benefit is conferred by a corporation on a shareholder of the corporation, … , then the amount or value of the benefit is to be included in computing the income of the shareholder, member or contemplated shareholder, as the case may be, for its taxation year that includes the time …

 

 

Part XIII, paragraph 214(3)(a):

 

214(3) Deemed payments -- For the purposes of this Part [XIII],

 

(a) where section 15 or subsection 56(2) would, if Part I were applicable, require an amount to be included in computing a taxpayer's income, that amount shall be deemed to have been paid to the taxpayer as a dividend from a corporation resident in Canada;

 

 

Part XIII, subsection 212(2):

 

212(2) Tax on dividends -- Every non-resident person shall pay an income tax of 25% on every amount that a corporation resident in Canada pays or credits, or is deemed by Part I … to pay or credit, to the non-resident person as, on account or in lieu of payment of, or in satisfaction of,

 

(a) a taxable dividend … or

 

(b) a capital dividend.

 

 

Part XIII, subsection 215(1):

 

215(1) Withholding and remittance of tax -- When a person pays, credits or provides, or is deemed to have paid, credited or provided, an amount on which an income tax is payable under this Part [XIII], … , the person shall, notwithstanding any agreement or law to the contrary, deduct or withhold from it the amount of the tax and forthwith remit that amount to the Receiver General on behalf of the non-resident person on account of the tax and shall submit with the remittance a statement in prescribed form.

 

 

Part XIII, subsection 215(6):

 

215(6) Liability for tax -- Where a person has failed to deduct or withhold any amount as required by this section from an amount paid or credited or deemed to have been paid or credited to a non-resident person, that person is liable to pay as tax under this Part on behalf of the non-resident person the whole of the amount that should have been deducted or withheld, and is entitled to deduct or withhold from any amount paid or credited by that person to the non-resident person or otherwise recover from the non-resident person any amount paid by that person as tax under this Part on behalf thereof.

 

 

Part XIII, paragraph 227(10)(d):

 

227(10) Assessment -- The Minister may at any time assess any amount payable under

…

(d) Part XIII by a person resident in Canada,

…

[368]   Article 9 of the Canada Luxembourg Treaty provides:

 

Article 9 Associated Enterprises 1. Where (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	Article 9 Entreprises associées 1. Lorsque a) une entreprise d’un État contractant participe directement ou indirectement à la direction, au contrôle ou au capital d’une entreprise de l’autre État contractant, ou que
(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,	b) les mêmes personnes participent directement ou indirectement à la direction, au contrôle ou au capital d’une entreprise d’un État contractant et d’une entreprise de l’autre État contractant,
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 income 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 income of that enterprise and taxed accordingly.	et que, dans l’un et l’autre cas, les deux entreprises sont, dans leurs relations commerciales ou financières, liées par des conditions convenues ou imposées, qui diffèrent de celles qui seraient convenues entre des entreprises indépendantes, les revenus qui, sans ces conditions, auraient été réalisés par l’une des entreprises mais n’ont pu l’être en fait à cause de ces conditions, peuvent être inclus dans les revenus de cette entreprise et imposés en conséquence.
2. Where a Contracting State includes in the income of an enterprise of that State — and taxes accordingly — income on which an enterprise of the other Contracting State has been charged to tax in that other State and the income so included is income which would have accrued to the enterprise of the first-mentioned State if the conditions made between the two enterprises had been those which would have been made between independent enterprises, then that other State shall make an appropriate adjustment to the amount of tax charged therein on that income. In determining such adjustment, due regard shall be had to the other provisions of this Convention and the competent authorities of the Contracting States shall if necessary consult each other.	2. Lorsqu’un État contractant inclut dans les revenus d’une entreprise de cet État — et impose en conséquence — des revenus sur lesquels une entreprise de l’autre État contractant a été imposée dans cet autre État, et que les revenus ainsi inclus sont des revenus qui auraient été réalisés par l’entreprise du premier État si les conditions convenues entre les deux entreprises avaient été celles qui auraient été convenues entre des entreprises indépendantes, l’autre État procède à un ajustement approprié du montant de l’impôt qui y a été perçu sur ces revenus. Pour déterminer cet ajustement, il est tenu compte des autres dispositions de la présente Convention et, si c’est nécessaire, les autorités compétentes des États contractants se consultent.
3. A Contracting State shall not change the income 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 income which would be subject to such change would, but for the conditions referred to in paragraph 1, have accrued to that enterprise.	3. Un État contractant ne rectifiera pas les revenus d’une entreprise dans les cas visés au paragraphe 1 après l’expiration des délais prévus par sa législation nationale et, en tout cas, après l’expiration de cinq ans à dater de la fin de l’année au cours de laquelle les revenus qui feraient l’objet d’une telle rectification auraient, sans les conditions visées au paragraphe 1, été réalisés par cette entreprise.
4. The provisions of paragraphs 2 and 3 shall not apply in the case of fraud or wilful default.	4. Les dispositions des paragraphes 2 et 3 ne s’appliquent pas en cas de fraude ou d’omission volontaire.

 

c) Positions of the Parties

 

[369]   It is the Appellant’s position that Article 9(3) of the Treaty applies and the assessment was barred by Article 9(3) as it was issued outside of the five-year period.

 

[370]   It is the Respondent’s position that, while CRA’s Part XIII assessment for McKesson Canada’s vicarious liability for an amount equal to the amount that should have been withheld and remitted to CRA by it when it paid MIH is the same as the amount of Canadian tax that would have been payable by MIH on its deemed dividend income, the standalone obligation of McKesson Canada under subsection 215(6) of the Act as a Canadian payor who fails to remit is distinct for Article 9 purposes from a change in the income of MIH for tax purposes resulting from the benefit conferred and the resulting deemed dividend.

 

[371]   The Respondent also argues that the description of income in Article 9 of the Treaty does not extend to deemed dividend income. It argues that a deemed dividend that accrues precisely because of the non-arms’ length relationship can not be considered to be income described in Article 9(1). The Crown argues that neither a benefit nor a deemed dividend could have accrued to MIH if the non-arm’s length conditions were removed from the RSA that is, if the Discount Rate had been the appropriate arm’s length discount rate. This would be the case only if the RSA Discount Rate were computed on arm’s length terms and the resulting higher purchase price was paid by MIH for the transferred receivables, or if less than all of the receivables portfolio had been transferred to reflect the RSA’s understatement of value. In either case, there would be no benefit conferred under the adjusted transactions with the result that there would be no deemed dividend and there could only be dividend income from McKesson Canada arising to MIH if McKesson Canada paid a dividend or conferred a benefit outside the RSA once adjusted to reflect arm’s length terms. The Crown argues that a real dividend could have accrued to MIH if the non-arm’s length conditions were removed from the RSA but only if McKesson Canada had separately declared a dividend for that particular given amount.

 

d) The Interpretation of Treaties

 

[372]   The Vienna Convention on the Law of Treaties[78] provides that a treaty is to be interpreted in good faith in accordance with the ordinary meaning to be given to the terms of the treaty in their context and in light of its object and purpose. It also authorizes regard to subsequent practice in the application of the treaty in certain circumstances and for certain purposes, as well as the use of other supplementary means of interpretation when the interpretation of the treaty otherwise leads to a result which is manifestly absurd or unreasonable.

 

[373]   In The Queen v. Crown Forest Industries Limited et al., 95 DTC 5389, the Supreme Court of Canada wrote: “In interpreting a treaty, the paramount goal is to find the meaning of the words in question. This process involves looking to the language used and the intentions of the parties.” The Court went on to quote approvingly from Addy J. in Gladden Estate v. The Queen, 85 DTC 5188, wherein he wrote at p. 5191:

 

Contrary to an ordinary taxing statute a tax treaty or convention must be given a liberal interpretation with a view to implementing the true intentions of the parties. A literal or legalistic interpretation must be avoided when the basic object of the treaty might be defeated or frustrated insofar as the particular item under consideration is concerned.

 

[374]   Both the Vienna Convention and the Supreme Court of Canada in Crown Forest confirm that “literalism has no role to play in the interpretation of treaties”: Coblentz v. The Queen, 96 DTC 6531 (FCA).

 

[375]   In Crown Forest the Supreme Court of Canada also held that, in ascertaining the purposes of a treaty article, a court may refer to extrinsic materials which form part of the legal context, including model conventions and official commentaries thereon, without the need to first find an ambiguity before turning to such materials.

 

[376]   The Preamble to the Canada-Luxembourg Treaty sets out its purposes of avoiding double taxation of income earned by a resident of one country from sources in the other country, and of preventing fiscal evasion. In Crown Forest the Supreme Court of Canada held that the purposes of the Canada-U.S. Treaty also included the promotion of international trade between the two countries and the mitigation of administrative complexities arising from having to comply with two uncoordinated taxation systems.

 

[377]   In The Queen v. Prévost Car Inc., 2009 FCA 57, 2009 DTC 5053, the Federal Court of Appeal affirmed the possible relevance of the OECD Commentaries to the OECD Model Convention, including commentaries subsequent to a particular treaty being entered into.

 

            e)  Analysis

 

[378]   Paragraph 1 is the primary transfer pricing adjustment paragraph of Article 9 of the Treaty. For purposes of McKesson Canada’s appeal, it provides that if either (a) MIH controls McKesson Canada or (b) McKesson U.S. participates directly or indirectly in the management or control of both MIH and McKesson Canada, and (c) the conditions of their financial or commercial relations differ from those conditions which would be made between independent parties, then: (d) any income which would have accrued to McKesson Canada but for those differing conditions but did not so accrue because of those conditions, may be included in McKesson Canada’s income and subject to Canadian tax.

[379]   Article 9(1) is in question in this appeal because it is referenced in paragraph 3 of Article 9. It can be noted that Article 9(1) clearly permits either state, Canada or Luxembourg, to tax either company if the preconditions are triggered. That is, in appropriate circumstances Canada is permitted by Article 9(1) to make a transfer pricing adjustment to MIH’s income subject to tax in Canada, such as if MIH carried on business in Canada in which it entered into non-arm’s length transactions on non-arm’s lengths terms with non-arm’s length parties. This makes obvious sense. An issue raised in this part of McKesson Canada’s appeal is whether Article 9(1) also addresses Canada indirectly adjusting the amount of deemed Canadian sourced dividend income of MIH resulting from its shareholder benefits or appropriations from McKesson Canada by virtue of the overstated Discount Rate in the RSA.

 

[380]   Paragraph 2 is the corresponding adjustment paragraph of Article 9. For purposes of McKesson Canada’s appeal, it provides that if Canada includes a transfer pricing adjustment amount in McKesson Canada’s income and MIH has already paid Luxembourg income tax on that amount, then Luxembourg shall make the appropriate corresponding adjustment to the Luxembourg income tax paid by MIH.

 

[381]   Article 9(2) is not at all in question in this appeal. It can be noted that Article 9(2) clearly only permits a state to adjust the tax paid to it by the enterprise of that state. This also makes obvious sense.

 

[382]   In this appeal there was no evidence that MIH paid any Luxembourg income tax on the income generated by it under the RSA from McKesson Canada’s receivables. The only evidence was Mr. Brennan’s handwritten note that some undescribed tax could be expected to be payable to Luxembourg as a result of the purchase and collection of McKesson Canada’s receivables under the RSA. This was not described as an income tax. There was also no evidence that any adjustment to MIH’s Luxembourg tax was needed to relieve any double tax, or whether such relief was either requested by MIH or granted by Luxembourg.

 

[383]   Paragraph 3 of Article 9 is the paragraph which imposes a five-year limitation period for making certain described transfer pricing adjustments. For purposes of McKesson Canada’s appeal of its Part XIII assessment for the amount it should have withheld from MIH and remitted to CRA in respect of MIH’s Part XIII Canadian tax liability on the benefit (or deemed dividend) of the overstated Discount Rate and resulting underpayment by MIH to McKesson Canada under the RSA for the transferred receivables, paragraph 3 provides that Canada shall not change the income of MIH in the circumstances referred to in paragraph 1 after a specific five-year period. That period is five years from the end of the year in which the income of MIH sought to be adjusted by Canada would have accrued to MIH but for the conditions referred to in Article 9(1).

 

[384]   Clearly Article 9(3) provides a maximum five-year limit (except in cases of wilful default or fraud)[79] for either state to make an Article 9(1) transfer pricing adjustment. It also clearly provides the same time limit on the other state having to make the corresponding adjustment on its counterparty under Article 9(2).

 

[385]   It is clear that Article 9(3) can apply to an assessment by either country of either the Canadian or Luxembourg party since it refers to the circumstances referred to in Article 9(1) which can so apply.

 

[386]   McKesson Canada argues that Article 9(3) also prevents Canada from assessing it under subsection 215(6) after March, 2008. In order for this argument to prevail, the following requirements of Article 9(3) (and by cross reference, Article 9(1)) must be met:

 

(i)      the subsection 215(6) assessment of McKesson Canada must be a change in the income of MIH.

 

(ii)     that adjustment of MIH’s income must be in the circumstances referred to in Article 9(1), namely:

 

(a)     MIH controls McKesson Canada or both MIH and McKesson Canada are indirectly managed or controlled by McKesson U.S.;

 

(b)     The conditions of the financial or commercial relations between MIH and McKesson Canada differ from the conditions which would have been made between independent parties;

 

(c)      The income adjustment is income which would have accrued to MIH, not McKesson Canada, but for those differing conditions in their financial and commercial relations; and

(d)     Canada sought to add the income adjustment to MIH’s income and taxed it accordingly.

 

(iii)    a period of five years must have passed since the end of the year in which the income of MIH sought to be changed would, but for the conditions which differ from what independent parties would agree to, have accrued to MIH.

 

[387]   In my view this argument can not prevail because Canada’s subsection 215(6) assessment of McKesson Canada does not satisfy all of these requirements.

 

[388]   Firstly, I question whether a subsection 215(6) vicarious assessment of a Canadian payor for failure to remit and withhold tax is a change by Canada of MIH’s income (requirement (i) above), or constitutes Canada seeking to add a transfer pricing adjustment amount to MIH’s income and tax that increased amount (requirement (ii)(d) above).

 

[389]   I am more inclined to see it as an enforcement and collection provision than a tax charging provision. Subsection 215(6) permits CRA to assess the Canadian payor an amount determined by reference to the amount it should have remitted to CRA but did not, which withholding amount is in turn determined by reference to the amount of Canadian tax that would have been payable by the non-resident payee. In the circumstances, however, I do not have to decide this point to dispose of this appeal.

 

[390]   It does, however, appear clear that an assessment of McKesson Canada for its failure to withhold and remit does not constitute Canada adding the transfer pricing income adjustment to MIH’s income and then taxing it accordingly (requirement (ii)(d) above). Adding it to MIH’s income and taxing it accordingly requires that Canada sought to tax MIH.

 

[391]   Secondly, these requirements are more clearly not met because the only transfer pricing adjustment in Article 9(1) is income which, but for the related party conditions, would have accrued to MIH under the RSA transactions (requirement (ii)(c) above). While the amount of MIH’s taxable benefit and deemed dividend may be the same as this transfer pricing adjustment, it is not an amount of income that, had the RSA had an arm’s length discount rate, would have accrued to MIH. On the contrary, the transfer pricing adjustment is income that but for the non-arm’s length terms and conditions would have accrued to McKesson Canada.

[392]   Had the RSA used an arm’s length discount rate and not the non-arm’s length Discount Rate actually used, the adjustment permitted by Article 9(1) can only be the additional McKesson Canada income. There would have been no excess benefit to, or appropriation by, MIH to be taxed as a deemed dividend, and there would not have been any actual dividend either unless a dividend was declared and paid by McKesson Canada to MIH which was also not the case. Clearly, Article 9(3) can not be read to apply to MIH’s deemed dividend arising from the non-arm’s length Discount Rate having been paid. For this reason alone, the taxpayer’s appeal can not succeed in respect of the Part XIII assessment.

 

[393]   Thirdly, this same fatal problem arises equally clearly yet again in respect of the Article 9(3) requirement (described in (iii) above) that a five-year limitation period can only begin to run from the end of the year in which the income of MIH sought to be changed would, but for the non-arm’s length Discount Rate used in the RSA, have accrued to MIH. Again, had an arm’s length Discount Rate been used in the RSA in McKesson Canada’s year ending March 31^st, 2003, the additional income would have accrued to McKesson Canada not MIH, as MIH would have paid McKesson Canada more for its receivables. Clearly, the “but for” wording of the Treaty requires the arm’s length conditions be substituted for the non-arm’s length conditions and, if this is done, the Article 9(1) and 9(3) adjusted income amounts can only be read as amounts that would have accrued to McKesson Canada, not MIH. This is a third independent reason why Article 9(3) can not relieve McKesson Canada from its liability under the Part XIII assessment for its failure to withhold and remit upon transferring its receivables to its non-resident parent for less than their value after agreeing to an excessive Discount Rate in the RSA.

 

[394]   If I may use the term primary transfer pricing adjustment to describe an Article 9(1) adjustment such as Canada’s addition of additional income to McKesson Canada, and the term corresponding transfer pricing adjustment to describe any corresponding downwards adjustment that may be made by a treaty partner to the counterparty’s income from non-arm’s length transactions by virtue of Article 9(2), I could only describe any taxation by the country making the primary transfer pricing adjustment of the excess amount of money wrongly appropriated by the counterparty in the other country, whether by way of deemed dividend or otherwise, as a secondary adjustment relating to the primary transfer pricing adjustment but not itself capable of being a primary transfer pricing adjustment described in Article 9(1) or thus by Article 9(3). Such secondary related adjustments can never meet the requirements of these paragraphs of Article 9 and therefore these shareholder benefits and appropriations of their subsidiary’s cash or valuable assets can not benefit from the provisions of these paragraphs.

 

[395]   This does not appear to be an inappropriate result when looked at from a Treaty purpose or a policy point of view. There was virtually no evidence of double taxation of the same income. At most, US$29,000 dollars of Luxembourg tax may have been payable on some basis by MIH as a result of the RSA. That was not on any shareholder benefit or appropriation or on any dividend or deemed dividend income. That amount is minuscule when compared with the millions of dollars of Canadian tax sought to be avoided. The double tax, if any, may have been the subject of a request by MIH for a corresponding adjustment from Luxembourg. I can assume that, if circumstances in Luxembourg warranted a corresponding adjustment, the McKesson Group’s Tax department would have applied for it.[80] Further, if McKesson Canada has a complaint that it has effectively had to pay MIH’s Canadian tax on MIH’s deemed dividend income arising from its shareholder benefit or appropriations from McKesson Canada, the Act gives McKesson Canada the right to seek indemnity from MIH. Further, given their common control within the McKesson Group, MIH might be expected to simply indemnify McKesson Canada without McKesson Canada pursuing MIH.[81]

 

          (f)      Conclusion re: Part XIII and the Treaty

 

[396]   In conclusion, the five year limitation period in Article 9(3) of the Treaty does not apply to the assessment of McKesson Canada’s vicarious liability for the amount of Part XIII tax payable by MIH which results from McKesson Canada’s failure to withhold and remit such amount. As shown above that is the result of the clear wording of Article 9(3), consistent with the overall context of Article 9, and consistent with the purposes of Treaty. Paragraph 227(10)(d) of the Act otherwise permits a subsection 215(6) assessment to be made at any time. For these reasons, the taxpayer’s appeal of its Part XIII assessment is also dismissed.

13.     Dismissal of Appeals

 

[397]   The taxpayer’s appeals are dismissed, with costs.[82]

 

 

       Signed at Edmonton, Alberta this 13^th day of December 2013.

 

 

"Patrick Boyle"

 

CITATION:                                      2013 TCC 404

 

COURT FILE NOS.:                         2008-2949(IT)G; 2008-3471(IT)G

 

STYLE OF CAUSE:                         MCKESSON CANADA CORPORATION AND HER MAJESTY THE QUEEN

 

PLACE OF HEARING:                    Toronto, Ontario

 

DATE OF HEARING:                      Oct. 17-20, 2011, Oct. 25-28, 2011, Oct. 31-Nov.

2, 2011, Nov. 15-18, 2011, Nov. 29-Dec. 2, 2011, Dec. 12-15, 2011, Jan. 16-20, 2012, Jan. 31-Feb. 3, 2012

 

REASONS FOR JUDGMENT BY:  The Honourable Justice Patrick Boyle

 

DATE OF JUDGMENT:                  December 13, 2013

 

APPEARANCES:

Counsel for the Appellant:	Paul B. Schabas, Ryder Gilliland, Jeffrey Trossman, Ilan Braude, Kaley Pulfer
Counsel for the Respondent:	Guy Laperrière Janie Payette Sylvain Ouimet Chantal Roberge

COUNSEL OF RECORD:

       For the Appellant:

                          Name:                     Paul B. Schabas, Ryder Gilliland, Jeffrey

                                                          Trossman, Ilan Braude, Kaley Pulfer

 

                            Firm:                     Blake, Cassels & Graydon LLP

                                                          Toronto, Ontario

 

       For the Respondent:                  William F. Pentney

                                                          Deputy Attorney General of Canada,

                                                          Ottawa, Canada