There are very few Canadian tax cases involving transfer pricing. As such, the Tax Court of Canada’s decision in McKesson Canada Corporation v The Queen[1] came as a welcome addition to the jurisprudence.

In McKesson, the Court rendered decisions on two appeals. The first and main appeal dealt with a primary adjustment on transactions between a Canadian Company (“McKesson Canada”) and related non-Canadian entities, and in particular, the sale of McKesson Canada’s receivables to its Luxembourg parent company at a discount. After analyzing several expert reports on non-arm’s length financial transactions, the Court held that the discount rate used by McKesson Canada and its Luxembourg parent company was higher than it should have been under the arm’s length principle.

The subject of this article is the related appeal which dealt with whether a secondary adjustment could be made beyond the five year time limit set out in Article 9 of the Canada-Luxembourg Income Tax Convention (the “Treaty”). In McKesson, an assessment for the transfer pricing primary adjustment was issued a few days before the five year limit expired; however, an assessment for the secondary adjustment was issued three weeks later. Justice Boyle upheld the secondary adjustment assessment finding that the five year limitation period in Article 9 did not apply, thereby allowing the Canada Revenue Agency (“CRA”) to assess McKesson Canada for its failure to withhold.

What is a Secondary Adjustment?

When a transfer pricing adjustment is made, CRA has determined that the arm’s length transfer price in respect of a transaction between non-arm’s length parties is different from the transfer price the parties originally chose for themselves and reported for tax purposes. The adjustment, usually referred to as the primary adjustment, is made pursuant to subsection 247(2) of the Income Tax Act (“Act”) and is generally an adjustment to the Canadian resident taxpayer’s understated revenues or overstated expenses. The primary adjustment addresses Part I taxes, but since value is typically shifted from a Canadian resident to a non-resident when the transfer price is incorrect, another adjustment is needed to ensure Canada collects the appropriate Part XIII withholding taxes. In the transfer pricing world, this is referred to as the secondary adjustment and is usually raised by CRA against the Canadian company as a failure to withhold[2] assessment.  

Prior to the Budget 2012 amendments to the Act, secondary adjustments were raised using traditional benefit conferral provisions[3] in the Act. The amendments of 2012 added provisions dealing specifically with secondary adjustmentsUnder the new provisions, notably subsection 247(12) of the Act, the amount of the primary adjustment results in a deemed dividend to the non-resident party, thereby invoking Part XIII withholding tax.[4] However, in McKesson the taxation year at issue was 2003, so the new provisions did not apply.

Treaty Time Limits in Article 9 of Canada’s Tax Treaties

Canada’s tax treaties often include a limitation period which prevents the tax administrations in both contracting states from raising transfer pricing adjustments beyond a certain period of time. This limitation period often varies between five and seven years and over-rides the domestic statutory limitation period of the contracting states[5].  The time limit in paragraph 3 of Article 9 of the Treaty reads as follows:

“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.”

As stated in paragraph 10 of the Commentary to Article 9 of the OECD Model Tax Convention, some states are not prepared for their correlative relief[6] obligations to last indefinitely. Consequently, those states, Canada included, often insist on including limitation periods (“treaty time limits”) in Article 9 of their treaties. Canada’s reluctance to provide open-ended correlative relief is partly attributable to its rules regarding the retention of tax records which would, in some circumstances,  make it difficult to verify the accuracy of foreign initiated transfer pricing adjustments for which the Canadian company’s records have been destroyed.[7] 

McKesson is the first Canadian jurisprudence dealing with both treaty time limits in Article 9 and secondary adjustments.

Justice Boyle’s Analysis of the Secondary Adjustment Assessment

Before considering Justice Boyle’s decision, the authors feel compelled to warn the readers that in the context of a typical transaction, the text of a treaty time limit rule and the general transfer pricing rule in paragraph 1 of Article 9 must be analyzed closely – one must proceed in super slow-motion in order to grasp the objective of the provisions. A typical transaction in a Canadian context would be where the enterprise being audited is a Canadian company and a primary adjustment is being made to the Canadian company that will create an income inclusion which would have accrued to the Canadian entity had the two related parties transacted in accordance with the arm’s length standard.  By way of contrast, the Court in McKessonwas being asked to interpret the interaction of these two Treaty provisions in the context of a secondary adjustment where CRA assessed Part XIII withholding tax pertaining to a divided deemed to have been received by the Luxembourg parent company. The two relevant provisions in the Treaty do not work well in the context of secondary adjustments, (applying these provisions in the context of secondary adjustments is like trying to force a square peg into a round hole).

In the context of a deemed dividend to the Luxembourg parent company, paragraph 3 of Article 9 provides that Canada shall not change the income of the Luxembourg entity in the circumstances referred to in paragraph 1 of Article 9 after the treaty time limit has passed. Consequently, the main issue considered by the Court was whether the description of income in paragraph 1 of Article 9 extended to deemed dividend income. The relevant portion of paragraph 1 of Article 9 reads as follows:

“…..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.” 

Justice Boyle looked at the wording of paragraph 1 of Article 9 and held that for the Article to apply in the context of the secondary adjustment (deemed dividend received by the Luxembourg parent company), three conditions must be met:

  1. the Part XIII assessment must be a change in the income of the Luxembourg parent company;
  2. the adjustment must meet the conditions of paragraph 1 of Article 9 including that
    1. the income adjustment be income which would have accrued to the Luxembourg parent company but for the non-arm’s length conditions, and
    2. Canada must have sought to tax the Luxembourg parent company’s income; and
  3. five years must have passed since the change in the income of the Luxembourg parent company.

Justice Boyle held that the second and third conditions listed above were not met. The amount at issue (the deemed dividend) was not income that would have accrued to the Luxembourg parent company but for the non-arm’s length conditions (i.e. the discount rate used by the parties). On the contrary, but for the non-arm’s length conditions, the amount at issue would have accrued to McKesson Canada from a higher sales price for the receivables (i.e. square peg – round hole).  Furthermore, because the amount at issue did not meet the second condition listed above, the five year time limit did not begin to run. Justice Boyle also noted in obiter that he was inclined to see Part XIII as more of an enforcement and collection provision than a tax charging provision, and he questioned whether the first condition listed above was met. However, he did not decide this point in disposing of the appeal.

Consequently, the assessment under Part XIII was considered to be on McKesson Canada, as a failure to withhold, and did not involve taxing the Luxembourg parent company in a manner contemplated by Article 9. These comments by Justice Boyle were consistent with the Respondent’s argument that “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.”[8]

The Court did not embark on a detailed analysis of the spirit and object of treaty time limits, but instead took a more literal approach to treaty interpretation. Nevertheless, the Court’s conclusion that paragraphs 1 and 3 of Article 9 do not apply to secondary adjustments can be supported by consideration of the purpose of Article 9. Paragraphs 8 and 9 of the Commentary to Article 9 of the OECD Model Treaty would seem to provide further support for concluding that the Article was not intended to apply to secondary adjustments.  For example, paragraph 8 of the Commentary states that “It is not the purpose of the paragraph to deal with what might be called “secondary adjustments”. 

One aspect of Justice Boyle’s analysis that the tax community should be particularly wary about is the statement that a failure to withhold assessment is “an enforcement and collection” mechanism and does not result in an “income” inclusion in the context of Article 9.  These comments, in the authors’ opinion, are a potential minefield and could have significant ramifications in the context of other treaty matters. This is touched upon at the end of this article.

Will there be Ramifications from the McKesson Decision on CRA Assessing Practices for Secondary Adjustments?

A situation that often arises during a CRA transfer pricing audit is that a treaty time limit for assessing a taxation year under audit will expire shortly, or has already expired. CRA auditors are generally fully aware of treaty time limits when conducting their transfer pricing audits and complete the audits and assess the primary and secondary adjustments within those treaty time limits. In our experience, once a particular taxation year is beyond the treaty time limit, CRA generally will not raise either the primary or secondary adjustment.  In the exceptional case where primary and secondary adjustments are raised beyond the treaty time limit, for example, where CRA mistakenly overlooked the treaty time limit[9], both the primary and secondary adjustments are usually overturned in a dispute resolution process (e.g. through the appeals process or a competent authority request under the Mutual Agreement Procedure (“MAP”) Article of the relevant treaty).

Is there a CRA Administrative Policy on Secondary Adjustments where a Primary Adjustment is not made because of an Article 9 Treaty Time Limit?

As alluded to above, in our experience CRA is generally fair and reasonable in its handling of the treaty time limits for transfer pricing cases. CRA generally does not pursue secondary adjustments if the primary adjustment is beyond the treaty time limit and is generally prepared to overturn the secondary adjustment in the dispute resolution process in those rare cases where the time limits were mistakenly overlooked by the auditors.  The rationale for this would appear to be one of administrative policy since, as can be seen from McKesson, CRA has the legal capacity to raise secondary adjustments beyond the treaty time limit. The authors are not aware of any written CRA administrative policy on this matter, and any informal policy could only be explained by CRA.  However, CRA may consider the following factors:

  • Raising a secondary adjustment where a primary adjustment will not be made still requires CRA auditors to undertake a full scale economic analysis of the transactions to determine the arm’s length transfer price and the amount of the potential benefit conferred. This analysis would require a significant investment of CRA resources.
  • In the event a secondary adjustment was raised by CRA and the primary adjustment was not raised, the related entities would still be eligible for dispute resolution recourse under the MAP Article of the relevant treaty.  Theoretically this could require a full scale bilateral negotiation between the competent authorities of both contracting states to come to an agreement on the arm’s length amount. This process would again result in a significant investment of resources of two government departments in an effort to resolve a secondary adjustment.
  • Canada’s treaty partners generally do not raise secondary adjustments where the primary adjustment is beyond the treaty time limit.
  • Questions would arise as to whether CRA’s repatriation policies[10] apply in circumstances where a secondary adjustment was raised but the primary adjustment was beyond the treaty time limit. Would the new repatriation policies in subsections 247(13) and (14) apply?[11] If these repatriation policies apply, this would seem to beg the question as to why CRA would invest such resources in raising the secondary adjustment by itself.   
  • Treaty time limits for primary adjustments are generally introduced into Canada’s tax treaties at Canada’s request. In principle, Canada’s policy rationale for insisting on these limits for primary adjustments should apply equally to secondary adjustments. That is, where Part XIII assessments are made by Canada, the treaty partners may be required to provide relief by way of foreign tax credits and may have treaty policy concerns regarding the provision of such relief for an open-ended period of time.

In McKesson, a primary adjustment was assessed by CRA within the treaty time limit in Article 9.  There was simply a timing issue in processing the secondary adjustment which caused it to be raised beyond the treaty time limit.  However, regardless of whether or not the primary adjustment was raised by CRA, the interpretive issues regarding the interaction of Article 9 treaty time limits and secondary adjustments are the same. Hopefully CRA will continue to use their good judgment in not pursuing secondary adjustments where primary adjustments are not raised due to a treaty time limit notwithstanding their legal ability to do so under McKesson.    

Impact of McKesson Decision on Treaties with a MAP Article Time Limit

The issue that will likely arise in the future is whether CRA will raise a secondary adjustment beyond a treaty time limit (where a primary adjustment is made on time) if the particular tax treaty has a similar time limit in the MAP Article (“MAP time limit”).  InMcKesson, the Treaty did not have a MAP time limit. However, it is not uncommon for MAP Articles of Canada’s tax treaties to have time limits similar to those of Article 9. A typical MAP time limit[12] is worded as follows:

A Contracting State shall not, after the expiry of the time limits provided in its domestic law and, in any case, after six years from the end of the taxable period in which the income concerned has accrued, increase the tax base of a resident of either of the Contracting States by including therein items of income which have also been charged to tax in the other Contracting State. This paragraph shall not apply in the case of fraud or wilful default.”

The Canadian MAP time limit seems to have originated as an after-thought to the Article 9 treaty time limit[13] to ensure that the taxpayer’s state of residence would not be required to provide correlative relief for permanent establishment income allocation adjustments made by the foreign jurisdiction unless those adjustments were carried out within a certain period of time.[14]  However, the wording of the MAP time limit has much broader application and actually may prevent a contracting state from assessing its residents, as opposed to preventing the assessment of income attributable to a Canadian permanent establishment of a resident of the other contracting state, in certain situations that would otherwise be assessable under domestic law.  

It is not clear whether CRA can raise a secondary adjustment in a fact situation similar toMcKesson if the relevant treaty has a MAP time limit. Until McKesson, many would have assumed that CRA could not raise any Part XIII assessment beyond the MAP time limit, even if the assessment was a failure to withhold assessment against the Canadian payer. However, the comments by Justice Boyle in McKesson, albeit in the form of obiter, seems to provide support for CRA to pursue these assessments if it chooses. A discussion of the various arguments and policy considerations that might come into play in determining whether a MAP time limit should prevent CRA from raising a failure to withhold Part XIII assessment is beyond the scope of this article, but McKesson has now brought this issue to the forefront.   

One would hope that CRA will not use the obiter in McKesson as a backdoor approach to raising secondary adjustments past MAP time limits. Situations do occasionally arise where a primary adjustment is made on time but a secondary adjustment cannot be processed within the MAP time limit. It is common for CRA to assess a taxation year close to the expiry of the time limit, and in such cases, the actual processing of the secondary adjustment is often delayed for weeks, taking those assessments outside the MAP time limit.

If CRA adopts the policy of ignoring MAP time limits on failure to withhold assessments on the basis of arguments similar to the obiter in McKesson, this could have significant ramifications that go far beyond transfer pricing secondary adjustments. For example, following the obiter in McKesson, CRA would seem to be in a position to ignore any MAP time limit in raising any Part XIII assessment since the assessment is generally by way of a failure to withhold assessment against the Canadian payer. Consequently, the tax community will be curious to see how CRA’s assessing policies and practices evolve on this issue, and perhaps more importantly, how the Canadian Competent Authority will handle failure to withhold assessments made beyond a MAP time limit if in fact CRA attempts to raise such assessments.