The recent Canadian Budget included a proposal to legislate against ‘‘foreign affiliate dumping’’, purported to be a means by which foreign-controlled Canadian companies reduce tax liability by either taking on debt or using surplus cash to acquire shares of non-Canadian members of the foreign-based group. The following article assesses the proposals. The recent Canadian Budget included a proposal to legislate against ‘‘foreign affiliate dumping’’, purported to be a means by which foreign-controlled Canadian companies reduce tax liability by either taking on debt or using surplus cash to acquire shares of non-Canadian members of the foreign-based group. The following article assesses the proposals as they were tabled by the Canadian government on March 29. The article was submitted for publication prior to an August 14 release of a revised set of proposals, which are referred to in the Postscript.

I. Overview

This article examines a current Canadian exercise in the ubiquitous and spreading arena of legislating what was once left to the courts, anti-avoidance law in this case rules related to certain transactions undertaken by foreign-owned/controlled Canadian corporations (and, in the view of the rule writers, undertaken because they are so controlled.)

The initiative, unveiled in the government’s March 29/12 Budget,1 and referred to as ‘‘debt dumping’’ or ‘‘foreign affiliate dumping’’, raises – as will be discussed below – two potential issues with antiavoidance law: (1) policy design defects2 and (2) implementation defects.3

The cycle and dynamic starts with the tax policymaker being concerned that a taxpayer is using a tax rule for purposes, or in a context, which differ from those intended by the legislator and which were not anticipated by the legislature and with the result that there is, in the tax policymaker’s view, inappropriate avoidance or deferral of tax, and it is that ‘‘mischief’’ to which the ‘‘anti-avoidance’’ rule is responding.4

That characterises the Canadian dumping proposals, which suffer from both policy design and implementation defects.  

II. The Canadian dumping issue

The Canadian dumping initiative stems from perceived mischief revolving around the use of two key provisions in Canadian law and the avoidance of a third.

One key rule is that a taxpayer can deduct, in computing overall taxable income – including that derived from Canadian business operations – interest on debt taken on to acquire shares in, inter alia, non-resident corporations.5 The tax policy rationale is to look at the potential overall economic and tax benefits for Canada stemming from the investment.6

The other key rule (and notably applicable even where the first rule applies) is Canada’s version of a participation exemption system for dividends from a foreign participation. This sees, under, inter alia, sections 90-95 and 113 of the Act, dividends received by a Canadian corporation from a non-resident corporation (NRC)7 that is a ‘‘foreign affiliate’’ (‘‘FA’’, a NRC in which the Canadian owns inter alia at least 10 percent of any class of its stock) exempt from tax, provided that the dividend stems from the FA’s ‘‘exempt surplus’’. That, in general, comprises active (or deemed active) business profits earned by a FA resident and operating in countries with which Canada has tax or information exchange agreements.8 Again, there are good policy reasons, grounded in a context, inter alia, of the Canadian owner of the FA having substantial economic interests in the FA, ones that potentially can produce both yield and investment/capital appreciation.

The 2008 Advisory Panel9 identified a mischief stemming from a perceived unintended use of those two rules which may be illustrated as follows.

  • Canco carries on profitable business operations in Canada and is owned by a foreign corporation (Forco) which also owns a subsidiary (Subco) in a third country that has a tax treaty with Canada and which requires funds to expand its business in that country.
  • Among the various ways in which those funds could be raised, it is decided that Canco will borrow from a third party, at X percent, and will make them available to Subco by way of investment in fixed value preferred shares, that carry a fixed cumulative dividend of X percent plus a few basis point (spread), to be issued to Canco by Subco.
  • That arrangement would seem to have no measurable non-tax economic effect on or benefit for Canco, at present or in the future, but it has a very measurable tax effect; the tax paid to Canada by Canco will be reduced. That is because the interest payments will reduce Canco’s taxable income base (derived from its profitable Canadian business) but the dividends will not increase the base because they will qualify for the ‘‘exempt surplus’’ (participation exemption) system.

That archetype ‘‘mischief’’ was characterised the Advisory Panel as ‘‘debt dumping’’.

But the Panel also suggested (without sufficient cause in this observer’s view) that the government consider whether the results are appropriate where the investment by Canco is in growth (common share) equity of a company such as Subco, an investment that could come about, for example, from a total sale by Forco of the shares of Subco to Canco with the price paid by a combination of interest bearing debt and an issue of shares.10 That would also, at least at inception, see Canadian tax leakage, but the potential future economic benefits for Canco (and tax benefits for Canada) create a dynamic and equation totally different than the archetype frozen pref scenario.

The Panel did not raise a concern about investment, (returning to the illustrative context) by Canco in Subco where no debt is taken on by Canco. In other words, Canco would use cash on hand. But the government’s March 29 Budget took the position that this may also be a ‘‘mischief’’ because it, inter alia, is avoiding Canada’s withholding tax on dividends paid to foreign shareholders.11

Finally, the government (unlike the Panel) sees potential mischief of this latter type in debt investment by foreign-controlled Canadian corporations in foreign members of the group.12  

III. The response

A. Overview

The foregoing points to the government seeing investment outside of Canada by a foreign controlled Canadian company as generally tax motivated (with, factually, narrow exceptions): either to generate interest deductions to reduce a pre-existing Canadian taxable income base or to access Canadian cash without paying dividend withholding tax. And the government does not see, even with respect to common share investment, compensating benefit to Canada.

The government’s basic approach, as set out in the March 29 Budget, is to establish punitive/preemptive type results for such conduct, unless narrow exceptions can be met. That reflects an objective that the conduct cease and the government has made clear that if it doesn’t the penalties will be increased.

The basic approach is to impose a tax as though a dividend had been paid, even though it has not been paid and even though the Canadian company will continue to be treated as the owner of the FA shares that have been acquired and even though an ultimate actual dividend will attract another round of withholding tax. Ostensibly that is pre-emptive and being quite capable of achieving the objective. But where there is a wholly-owning parent, resident in a treaty country with a 5 percent dividend rate, rather the 25 percent statutory rate, that cost may be quickly recovered by tax savings from interest charges.

It is puzzling why one obvious response was not chosen, namely to make dividends on impugned investments ineligible for the exempt surplus rules. If that approach were taken with regard to fixed value prefs, the one area where the mischief is clear, (and if for this purpose the cumulative dividend were required to be accrued if not paid), there would no longer be any tax benefit from the debt dump and that would totally shut down that most blatant and obvious of the various types of transactions that concern the government.13

It is true that such approach for prefs would beg the question for common share investment. But the answer to that is simple. There should be no new rules, at all, for such investment.

B. Specifics

The new penalty regime – mainly contained in proposed new section 212.3 of the Act – consists of five elements:

  • the conditions for its application;
  • the exception to its application;
  • the type of investment to which it will apply;
  • the consequences of its application; and
  • a few subsidiary rules to give effect to the those first four primary elements.

The conditions for its application (stated in section 212.3(1)) are threefold:

  1. there is an ‘‘investment in a (pre-existing or new) FA;  
  2. the investor is a ‘‘corporation resident in Canada’’ (a CRIC that is controlled by a non-resident corporation; and  
  3. a ‘‘bona fide purpose’’ exception is not met. (The third element – the exception – is dealt with in the next section, below).  

The key point to note is that the regime only applies where there is a foreign corporate controller of the investing Canadian corporation and the definition of the term ‘‘investment’’ shows that the regime is not limited to the archetype mischief, investment in preferred shares. It extends to common shares and to loans to or other debt instruments of foreign affiliates. 14

Where the regime applies there will be a penalty imposed, either at the point the investment is made or, potentially at both that time and at a later time.  

Where the Canadian corporation pays for the investment by transferring property (other than its own shares) or by assuming a debt obligation, in partial or total consideration of the investment, such part of the payment is to be treated as if it were a dividend paid by the corporation to its controlling foreign corporate shareholder,15 that triggers the 25 percent WHT (subject to treaty rate reduction).

Where the Canadian corporation pays, in whole or in part, for the investment by issuing shares to the seller to it, (say, the ‘‘Forco’’ in the earlier example) a rule entailing three elements is triggered.16 Firstly, the reference in the rule to the ‘‘paid up capital’’ of the shares invokes a definition in section 89(1) of the Act which starts off by treating the PUC as the legal corporate share capital created under corporate law and then provides for certain adjustments. Secondly, under section 84, the PUC of shares of a Canadian resident corporation sets the maximum amount that the corporation can distribute (return) to a shareholder, upon a legal capital reduction or share redemption or buyback, without triggering a deemed dividend.17 Thirdly in light of those two rules the prime effect of the proposed rule – to deny any addition to PUC for shares issued upon the making of an impugned investment – is that any attempt at a future time to repay the company’s shareholders’ for the investment, by returning capital, will trigger a dividend.

As well the PUC of shares owned by non-resident shareholders governs the thin capitalisation rules (see above) so that the denial of PUC increase will limit the amount of deductible interest the CRIC can pay to foreign shareholders. In particular, having regard to the illustration above where Forco sells all of Subco to Canco, under current law a complying split of the consideration between debt and equity would see interest paid on the debt being deductible to Canco. But the new paragraph (b) would operate to oust that deductibility. That is the immediate impact of paragraph (b).

Both parts of the ‘‘consequences’’ rule raise design and implementation defects. Some have already been noted and others are discussed below.

But three points should be reiterated or noted at this juncture. Firstly, there is the basic issue of the excessive ambit of the rule and, in particular, its extension to investment in common shares. Secondly, there is the inherent and inappropriate inconsistency – productive of double tax – of not treating the deemed dividend as also distributing the investment to the Forco (with ‘‘Canco’’ holding it as agent for Forco). Thirdly, there is the excessiveness and uncertainty of applicable treaty rate relief with respect to the deemed dividend, where the Forco is not a wholly-owning shareholder. Consider for example a publicly-traded Canco where 51 percent of the shares are owned by the Forco and 49 percent by Canadian residents.

C. The limited exception

The third condition for application of the regime (section 212.(3)(c)) is that ‘‘the investment may not reasonably be considered to have been made by the CRIC, instead of being made or retained by the parent or another non-resident person that does not deal at arm’s length with the parent, primarily for bona fide purposes other than to obtain a tax benefit (as defined in subsection 245(1))’’.

That is a difficult notion to deal with. The proposals seek to provide assistance/guidance. In particular section 212.3(5) spells out seven factors that ‘‘are to be given primary consideration’’ in ‘‘ determining whether paragraph (1)(c) applies’’.  

But those factors – focusing on whether there is greater business and management nexus to Canco as opposed to the foreign members18 – will generally be bound up in such subjectivity as to render the exception unreliable. Just consider the feasibility of correlating the notion in section 212.3(5)(a), that the FA’s business is more ‘‘closely connected’’ to the CRIC than any foreign member of the group, or the notions in paragraph (c) through (g) of identifying who is running the CRIC with the complexity and fluidity/ constant flux and change in multinationals of any appreciable size. Apart from being daunting, can it ever be feasible, even when there are the purest of motives?

This approach reflects the government’s underlying view that tax benefits are generally the only reason a foreign-controlled Canadian corporation would be the investor, rather than the parent or other foreign sister in a foreign member of the group.19 Therefore, the rule is written to reflect the notion, stated in the Budget, pp 6–49, that taxpayers will have the onus of showing that the ‘‘...investment . . . belongs in the Canadian subsidiary more than in any other entity in the foreign parent’s group’’.

At the end of the day this approach would put relevant multinationals in a straitjacket and the concerns were perhaps best summed up in a few lines by Tax Executive Institute (TEI), which wrote in its May/June submission to Finance about the ‘‘excessive breadth’’ of the proposal, the ‘‘subjective and unworkable tasks’’ to be excluded and that ‘‘the proposal will substantially undermine the attractiveness of Canada as a destination for foreign investment’’.20

IV. Concluding comments and observations

As already indicated there are fundamental flaws in these proposals, some of a design type and some of an implementation type and some bridging the two. For example how should the two following hypotheticals, arising from the illustration used earlier, be characterised?

Suppose Canco purchased Subco from Forco on Day One for CAD 100m cash. There would be a deemed dividend at a Canadian tax cost of at least 5 percent CAD 5m of tax. Suppose on Day Two the parties change their plans and the shares of Subco are distributed to Forco as a dividend in kind. Here there is an actual dividend and another 5 percent. The proposals provide no relief for the double tax.

Or, suppose the sale, on Day One, sees Canco satisfy the price by issuing a note of CAD 60m and shares of CAD 40m and a day later the transaction is reversed with Forco satisfying the price by tendering back both the note and the shares. Other than that a deemed dividend of CAD 60m arose on Day One and of CAD 40m on Day Two, the parties are in the same position as before Day One: pay the government at least 5 percent.

Separately there is an overriding concern expressed by some observers that the proposals will fundamentally change Canada’s position in the international business community.21 This in part stems from the manner in which the rules will inhibit standard/ normal treasury activities that take place daily in MNEs.

Then there are concerns over the position of publicly traded companies. The difficulty is that there is a total conceptual disconnect between the premise of the dumping proposals, that the Canadian subsidiary is a puppet, acting at the command of its foreign parent, and the reality that publicly-traded companies are forced by a combination of market forces and securities law to act independent of any shareholder, including a controlling shareholder. Therefore there should be a complete carve out from the proposals for such companies.

Furthermore, the proposals do not correlate their ambit as to whether or not the CRIC has Canadian operations and a taxable income base to protect.22 That is particularly unfortunate in the resource sector where players worldwide choose Canada as a base for international activities and often emerge with a network of Canadian companies (included TSX or TVX listed companies) which have no or virtually no Canadian operations or tax base.23

A review of current developments in other countries indicate few (but see France and Spain) have adopted or are adopting rules of the type that Canada has proposed. 24

Finally, although submissions to the government on these proposals seem to uniformly condemn their design excessiveness (apart from the implementation defects that harbour the risk of premature and double tax and applicability to obviously unintended situations), only two of those examined by this writer (that by Deloitte and TEI) would clearly rein them in by totally excluding common share investment from their ambit.25 Time will tell whether the government heeds their views and recommendations.

In summary, Canada’s dumping proposals are beset with both fundamental design and implementation defects and it is hoped the Department of Finance will, as recommended by several observers, pull the proposals for further consultation and study.


A revised draft of the proposals, issued ( after the report above was written) on August 14 (Department of Finance release 2012-89 and related documents) , does not address the fundamental policy design defect discussed above – namely the potential applicability of the proposals to investment in common shares of foreign affiliates. In this writer’s view that (as well as a proposed extension of the rules to investment in certain Canadian companies that own foreign affiliates) overshadows a number of steps taken to ameliorate some of the most problematic of the implementation defects in the original draft. It remains to be seen if submissions requested by the government by September 13 lead to any material change in the final legislation. The overall matter will be dealt with in a sequel to this report once final legislation has been tabled.