GAAR: KEEPING US OFF-BALANCE?

In an episode of Corner Gas, the main character, Brent, (who is single, has no children and is 40 years old) finds himself overwhelmed by a hot-wheels-tossing 6 year old boy and besieged under a kitchen table while babysitting for a friend.  Brent enlists the help of his mother, Emma, who quickly gains control of the situation and explains her two-pronged philosophy for disciplining kids:

Use their whole name.  They hardly ever hear all three of their names so it kind of puts them off balance [mentally].  Second, give them an open ended threat, nothing specific, that way their fertile little imaginations fill in the gap with the worst thing they can think of.  A child’s imagination is their strength and their weakness.”

To watch this portion of the episode“Oh Baby,” start at 3:38 of the linked video.

Emma’s philosophy made me think of the general anti-avoidance rule (the “GAAR”).  Until the introduction of the GAAR in 1988, one of the greater certainties in tax, while not absolute, was the Duke of Westminster principle that taxpayers may order their affairs so as to minimize the amount of tax payable.  The GAAR introduced significant uncertainty and great apprehension as to whether taxpayers could seek to minimize tax.  Initially, the GAAR was feared as a “big stick” to reign-in taxpayers and tax planners alike.  As the GAAR cases worked their way through the courts, the apprehension may have diminished, but uncertainty remained.

The decision from the Supreme Court of Canada (“Supreme Court”) in Canada Trustco Mortgage Co. v. The Queen  2005 SCC 54 provided a welcome framework in applying the GAAR and restored a measure of certainty.  Particular comfort was gleaned from the following words of the Supreme Court:

“The GAAR was enacted as a provision of last resort in order to address abusive tax avoidance, it was not intended to introduce uncertainty in tax planning.”

With the Supreme Court’s split decision in Earl Lipson v. The Queen 2009 SCC 1, many commentators have expressed concern that Lipson introduced new and improved uncertainty.  (We discussed the facts of Lipson in our January 9, 2009 blog.)  Consider the view of one dissenting judge:

“The approbation by the Court of the Minister's resort to vague generalities or "overriding policy" would only increase the element of uncertainty in tax planning that Canada Trustco sought to avoid.” 

There are fears that the “big stick” is back and even whispers of a smell test.

Lehigh


On May 17, 2010, the “certainty” of the GAAR took another twist when the Federal Court of Appeal (“Federal Court”) allowed the taxpayer’s appeal in Lehigh Cement Limited v. The Queen, 2010 FCA 124, reversing the decision of the Tax Court of Canada (the “Tax Court”).  The Tax Court had applied the GAAR to a series of transactions that resulted in the avoidance of withholding tax under subparagraph 212(1)(b)(vii) of the Income Tax Act (the “Act”), which, in general terms, exempted interest from withholding tax when paid to an arm’s length party on debt that under no circumstances was 25% of the principal required to be repaid within 5 years of the date of issue of the debt ( the “5/25 Exemption”).  Canada has since eliminated withholding tax on all arm’s length interest (other than participating interest payments).

Misuse or Doubt?

The Crown’s primary argument in applying the GAAR in Lehigh was that a non-resident person is not entitled to benefit from the 5/25 Exemption where the right to receive interest is split from the right to receive the principal amount because the transaction did not result in Lehigh “accessing funds in an international capital market.”  This phrase is an excerpt from a 1975 Department of Finance budget paper that first proposed the 5/25 Exemption.

The Federal Court rejected the Crown’s invitation to conclude that entitlement to the exemption is “subject to a condition necessarily implied by the existence of a fiscal policy, evidenced only by a sentence in a 1975 budget paper that is said to explain why the exemption was enacted.”  The Federal Court emphasized: 
 

“Most importantly, if there is any doubt as to whether the transaction in issue results in a misuse […], Lehigh is entitled to the benefit of that doubt.”
 

Although the reversal of Lehigh itself by the Federal Court may be viewed as contributing to the uncertainty surrounding the GAAR, the message may be broader than that.  Arguably, Lehigh restores a measure of certainty to the principle that when there is any doubt as to whether there is a misuse or abuse of the provisions of the Act when a taxpayer is arranging his or her affairs to minimize tax, the taxpayer should be entitled to the benefit of the doubt.  It also inspires confidence that, as questioned by the Supreme Court in Canada Trustco, the Federal Court is unwilling to formulate taxation policies that are not grounded in the provisions of the Act and certainly will not apply ungrounded policies to override the specific provisions of the Act.  So it would seem that certainty can be found in doubt.
 

Quebec's Take on Fighting Aggressive Tax Planning

Why would we be writing a blog on future legislative amendments to the Quebec Taxation Act when we are Alberta tax advisors and most of our clients are in Western Canada? We believe the new Quebec proposals that are the subject of this blog may have national significance if the federal or provincial governments ever follow suit. Over the years, Quebec has been the subject of certain aggressive tax planning. It appears that Quebec has been seething over these aggressive transactions ever since and has finally come out with a heavy handed proverbial slap to the taxpayer's face.

On January 30, 2009 the Quebec Minister of Finance released a working paper entitled “Aggressive Tax Planning”. The purpose of the paper was to expose what the Quebec Minister of Finance considered to be aggressive tax planning (“ATP”) and actions being contemplated to curb ATP. Interested persons were asked to provide their comments on the Minister’s proposed actions up until April 1, 2009. The result of this consultation process was announced on October 15, 2009 with an information bulletin laying out Quebec’s initiative to combat ATP. While some of these measures are not as far reaching as feared, some of them have quite an impact on how tax planning advisory services may be provided in Quebec.

The following is a summary of some of the more important legislation to be enacted:

1. Mandatory disclosure transactions: The Quebec Minister of Finance has identified circumstantial factors relating to taxpayers and their advisors that are likely, in their opinion, to lead to ATP. These are confidential transactions and transactions with conditional remuneration. A taxpayer will have to disclose transactions in respect to which the taxpayer has entered into a confidentiality agreement. The de minimis threshold under which a transaction need not be disclosed is a transaction with a tax benefit of less than $25,000 or an impact on income of less than $100,000.

Taxpayers will also be required to disclose transactions for which the tax advisor’s remuneration is conditional on obtaining a tax benefit, is refundable if the expected tax benefit does not materialize or is paid to the advisor only after the expiry of the limitation period applicable to the taxation year the transaction was undertaken.

Disclosure will have to be made in prescribed form under separate cover, by registered mail or electronically, on the due date for filing the taxpayer’s tax return. Failure to disclose will result in a $10,000 penalty with an additional $1,000 per day as of the second day, to a maximum of $100,000. Furthermore, failure to file a disclosure will result in the suspension of the limitation period relating to the undisclosed transaction. A defense of due diligence and the voluntary disclosure program will apply to the new disclosure rules.

2. General Anti-Avoidance Rule: More worrisome than the above mandatory disclosure requirements is the penalty provisions being added to the Quebec GAAR. Where the Quebec GAAR applies to a transaction, the normal limitation period will be extended by three years. This same approach was adopted by Alberta, effective December 2, 2008. In addition, subject to a due diligence defense, a penalty of 25% of the amount of the tax benefit will apply if GAAR applies to the transaction. Furthermore, when GAAR applies, the promoter of the transaction will incur a 12.5% penalty on all amounts of consideration received by the promoter, subject to due diligence. None of the foregoing extended limitation period or penalties will apply if the transaction was disclosed in the taxpayer’s return of income in prescribed form.

The new legislation will be effective after the day of publication of the information bulletin (i.e. October 16, 2009) for transactions carried out on or after the date of publication of the information bulletin, but not for series of transactions that began before this date. As you can see from the foregoing, the Quebec Minister of Finance has decided to get “tough” on what they perceive as being abusive tax avoidance. It remains to be seen what impact these provisions will have on business in Quebec and on tax advisors. It also remains to be seen whether the Federal Government, or other provinces, will follow Quebec’s lead and legislate in the same direction. This is yet another clear sign that tax authorities and related organizations (including the OECD) are taking tougher positions against tax avoidance. Stay tuned …

Barbados Spousal Trusts - Tax Avoidance

This blog is lengthier than usual given the importance of this topic. Antle v. The Queen, 2009 TCC 465, is an interesting new case released on September 18, 2009. It involved the use of a Barbados Spousal Trust to avoid Canadian capital gains tax on the sale of shares.

In Antle, the taxpayer husband rolled shares with inherent gains to a Barbados Spousal Trust. The Barbados Trust sold the shares to the beneficiary wife, who then sold them to a third party purchaser. The sale proceeds from the third party purchaser were used by the wife to pay off the Trust. The Trust made a tax-free distribution the beneficiary-wife. The trust was then immediately dissolved. No tax was payable, either in Canada or Barbados. (Barbados does not impose tax on capital gain, which is where the gain arose.) Justice Miller, writing for the Tax Court of Canada, found that the Barbados Spousal Trust was not validly constituted, and, even if it had been, GAAR would apply to deny any tax benefits to the taxpayer husband.

Although tax practitioners will likely be disappointed by the result of this case, the news is not altogether bad. In our view, the case significantly advances the discussion on what might constitute “abuse” under GAAR. For example, the distinction Justice Miller draws between “unintended loopholes” and “policy” is fairly novel, as is the notion that one cannot necessarily infer policy from an omission by Parliament to enact anti-avoidance rules. Whether or not one agrees with the ratio of Justice Miller in this decision, cases such as these add depth to the GAAR discourse and represent a step in the direction of having a GAAR which is more easily applied and predicted.

Furthermore, Justice Miller’s approach was laudable in that his conclusion on abuse was carefully substantiated. He stopped short of applying GAAR to the Barbados Trust, because going down that route would be “unreceptive to analysis”. On this point, he writes:

To suggest one can cobble together some policy underlying the interplay between Article XIV of the Treaty and the Income Tax Act provisions at issue is to provide fodder perhaps for academics, but it is, with all due respect to the drafters of both, an attempt to create an object, spirit and purpose –policy if you will – where none was contemplated. Certainly, no such policy was pointed out to me. Article XIV is what it is. [our emphasis]

Justice Miller thus appears to be introducing the notion that, in some cases, no policy is contemplated beyond the text of a provision itself. This represents a welcome development for tax planners. Indeed, if certain provisions lack an underlying policy beyond the text of the provision itself, then there can be no abuse of such provisions where the letter of the law is met.
Of less appeal to tax practitioners is Justice Miller’s reliance on a broadly worded, “overarching” policy of the OECD that tax conventions - generally - are to prevent tax avoidance and evasion. Whether this violates the prescriptions of Canada Trustco, 2005 SCC 54, namely, that the judiciary should not utilize an “overarching policy” to “override the wording of the provisions of the Income Tax Act” remains to be seen.

A detailed summary of the decision, with important passages in bold, follows.

Validity of Trust

The Court examined whether the Barbados Trust met the three certainties, namely: certainty of intention; certainty of subject matter; and certainty of objects. The Trust failed the first test. The appropriate intention to create a Trust was found to be lacking. According to the Court, a determination of intention cannot be based on the language of the Trust Deed alone; rather the intention to create a Trust must also be evidenced in the conduct of the parties. In this case, basing itself on a number of unfavorable facts, the Court found that the taxpayer never intended to lose control of the shares in favor of the Trustee, nor did the taxpayer fully appreciate the significance of settling a discretionary Trust. The Trust was only used as a conduit to avoid tax.

The Trust also failed the second test—certainty of subject matter. Although the Trust Deed contained a statement that the subject matter of the Trust was the shares, title to those shares was never in fact transferred to the Trustee. Indeed, at all times, the share certificates were held by a third person (the person from whom the taxpayer had originally acquired the shares) until the certificates were delivered directly to the ultimate purchaser. At no point were the certificates ever delivered to the Trustee, as required under the relevant provincial legislation.

Given the Barbados Spousal Trust failed the first two certainties, it was not found to be validly constituted. The judge concluded:

“(…) there is no properly constituted Trust: the Trust never came into existence. This conclusion emphasizes how important it is, in implementing strategies with no purpose other than the avoidance of tax, that meticulous and scrupulous regard be had to timing and execution. Backdating of documents, fuzzy intentions, lack of transfer documents, lack of discretion, lack of commercial purpose, delivery o0f signed documents distributing capital from the trust prior to its purported settlement, all frankly miss the mark—by a long shot. (…) In short, it is not enough to have a brilliant strategy, you must have brilliant execution.” [our emphasis]

Trust a Sham?

Although the Trust was found not to be valid, Justice Miller nonetheless turned his mind to whether the Trust (if it had been validly constituted) was a sham. The notion of sham requires an element of deceit which generally manifests itself by a misrepresentation by the parties of the actual transaction taking place between them. In the case of a trust, a finding of sham requires that both the settlor and the trustee share in the deceit. In this case, the taxpayer never intended the Trustee to have control of the shares. Moreover, the Trustee took directions from the taxpayer and failed to show any exercise of discretion. However, according to Justice Miller, such facts did not amount to a finding of intentional deceit as required by the doctrine of sham. Interestingly, Justice Miller agreed with the appellant’s submission that a tax motivation does not create a sham. Moreover, he agreed that the retention of power by a settlor to direct or influence trust investments does not result in sham.

GAAR

In applying GAAR, the Court quickly concluded there had been a tax benefit and an avoidance transaction. The Court then spent the next 15 pages considering whether the Barbados Spousal Trust strategy had been abusive within the meaning of subsection 245(4). As mandated by the Supreme Court in Canada Trustco, the Court purported to apply the unified textual, contextual and purposive approach in finding the object and spirit of the provisions of the Act or Treaty relied upon by the taxpayer to obtain the tax benefit. The Court specifically considered the object and spirit of the provisions relied upon by the taxpayer - section 73, paragraph 94(1)(c), paragraph 110(1)(f) of the Act, and Article XIV(4) of the Treaty - but also the provisions circumvented by the taxpayer, being the attribution rules in sections 74.1 to 74.5.

According to Justice Miller, the policy underlying subsection 73(1) of the Act is to recognize that spouses are a single “economic unit” by allowing a tax-deferred rollover of property between them. Viewed in the context of the spousal attribution rules contained in subsection 74.2(1) of the Act, Justice Miller found that the object, spirit and purpose of the rollover/attribution regime is to ensure that gains are taxed when property leaves the marital unit.

Justice Miller went on to consider the policy underlying paragraph 94(1)(c). (Paragraph 94(1)(c) was relevant because it deemed the Barbados Spousal Trust to be resident in Canada, which allowed the taxpayers to take advantage of the subsection 73(1) rollover, as that provision read at the time). Basing himself on Finance Technical Notes, he noted that paragraph 94(1)(c) is an anti-avoidance rule designed to prevent Canadian residents from deferring or avoiding Canadian tax by holding property in a non-resident Trust established for the benefit of Canadian resident beneficiaries.

The Barbados Spousal Trust then claimed the exemption in Article XIV(4) of the Canada-Barbados Treaty, pursuant to which the share sale would only be taxable where the trust was resident (Barbados). Paragraph 110(1)(f) of the Canadian Act further clarifies that amounts exempt from tax in Canada because of a tax treaty are to be deducted from a taxpayer’s income in the year. The Court noted that the purpose of Article XIV(4) of the Canada-Barbados Treaty and paragraph 110(1)(f) of the Act was to exempt from Canadian tax gains realized by Barbados residents on the disposition of capital property. However, the Court went on to look for a broader purpose of what tax treaties are generally intended to accomplish. It noted that, according to OECD Commentary, the purpose of tax treaties is to prevent tax avoidance and evasion. Importantly, the Court quoted the following examples of abuse from the OECD Commentary:

(…) for example, if a person (whether or not resident of the Contracting State) acts through a legal entity created in a State essentially to obtain treaty benefits that would not be available directly. Another case would be an individual who has in a contracting State both his permanent home and all his economic interests, including a substantial shareholding in a company of that State, and who, essentially in order to sell the shares and escape taxation in that State on the capital gains from the alienation (…), transfers his permanent home to the other Contracting State, where such gains are subject to little or no tax. [our emphasis]

In light of the above, the Court applied GAAR to Mr. Antle, the Canadian resident taxpayer. The Court stated in this regard: 

(…) by relying on paragraph 94(1)(c) to deem the Trust to be a Canadian resident to take advantage of the subsection 73(1) rollover, and then escaping Canadian tax liability by invoking paragraph 110(1)(f), (…) Mr Antle has blatantly frustrated the object, spirit and purpose of the rollover/attribution regime. (…)

It is an outcome that subsection 73(1), the attribution rules and paragraph 94(1)(c) specifically sought to prevent; that is, the marital unit cannot escape liability by using an offshore trust. (…)

Further, the outcome also defeats the underlying rationale of these particular provisions and indeed Canada’s policy of taxing capital gains generally. Canadian residents and deemed residents are to be taxed on their capital gains in Canada. Rules to capture the gain on disposition of capital property by the marital unit, in keeping with the Canadian policy, are rendered meaningless simply by finding a willing Barbados Trustee.

However, the Court would not go as far as to apply GAAR to the Barbados Trust itself. In this regard, the Court writes:

To apply GAAR to the Trust itself would require identifying the spirit, object and purpose of the Treaty provision (Article XIV) or to identify the policy of that Treaty provision together with the relevant Income Tax Act provisions read as a whole. The former analysis is not difficult: for the most part gains on disposition of movable property are to be taxed in the alienator’s country of residence. There is no further object or purpose to be found – it is clear from the provision itself. The latter analysis is not so straightforward. Indeed, it is, I suggest unreceptive to analysis. To suggest one can cobble together some policy underlying the interplay between Article XIV of the Treaty and the Income Tax Act provisions at issue is to provide fodder perhaps for academics, but it is, with all due respect to the drafters of both, an attempt to create an object, spirit and purpose –policy if you will – where none was contemplated. Certainly, no such policy was pointed out to me. Article XIV is what it is.

Finally, the Court went on to deal with a number of specific arguments advanced by the taxpayer as to why the plan was not abusive (all of which were rejected by the Court). The appellant argued, for example, that if former section 94 had applied, the Trust would have been taxable with no Treaty exemption. Under the new rules, the Trust is still taxable under paragraph 94(1)(c), but can claim the Treaty exemption in article XIV. While the Court agreed that this is how the rules operate, the legislative amendments could not used to infer that the Barbados Spousal Trust strategy was not abusive. According to the Court, the legislative amendments together with the Treaty exemption resulted in an “unintended loophole”, and that “loopholes are not policy makers”.

The appellant argued there was no abuse in this situation because of what the Canada-Barbados Treaty does not say. Namely, the Treaty puts no limits on who can create trusts, contains no denial-of-benefits clauses, limitation-of-benefits clauses, or anti-avoidance rules that would catch such transactions. Each of these arguments was rejected by the Court for fundamentally the same reason: such omissions do not lead to the inference that there is a policy in the Treaty allowing the Barbados Spousal Trust Strategy. The appellant’s argument that the Canadian government must have known that Canadian taxpayers could engage in such strategies because the Canada-Barbados Treaty invited them to do so was rejected by the Court. Rather, the Court was of the opinion that the Canadian government never contemplated that tax planners would come up with such tax avoidance plans.