5 May 2021 IFA Finance Roundtable

This provides summaries of some of the oral remarks provided at the IFA (Canadian Chapter) Finance Roundtable held online on 5 May 2021, along with brief summaries of questions posed. It also summarizes comments made by Phil Halvorson in presenting illustrative examples, which he had generally discussed with Shawn Porter. The panelists were:

Shawn Porter, Associate Assistant Deputy Minister, Tax Policy Branch, Finance Canada
Trevor McGowan, Director General, Tax Legislation Division, Finance Canada
Tim Barett, Thorsteinssons
Phil Halvorson, EY

Reportable and notifiable transactions

Tim Barrett – Budget 2021 includes a number of measures to assist officials in administering the Act, including the ability to require oral interviews while conducting an audit, and disclosure of uncertain tax positions for certain corporate taxpayers.

There are two measures on reportable transactions. The first is an amendment to the existing reportable transaction rules, and the other is a set of newly proposed “notifiable” transaction rules. Both appear to be inspired by the BEPS 12 recommendations.

What is the general distinction between “reportable” and “notifiable” transaction rules?

Trevor McGowan – The measures are probably inspired by the OECD BEPS project and Action 12, even going so far as lifting a lot of the language from Action 12, especially on hallmarks.

You asked about the differences between reportable and notifiable transactions. The first difference is that reportable transactions are tweaked version of what is currently in the Act, whereas notifiable transactions are a whole new set of rules.

Reportable transactions are referred to in the OECD documents as relating to “generic hallmarks,” and the notifiable transactions relate to “specific hallmarks. BEPS Action 12 states (at chapter 2, p. 39):

Hallmarks act as tools to identify the features of schemes that tax administrations are interested in. They are generally divided into two categories: generic and specific hallmarks. Generic hallmarks target features that are common to promoted schemes, such as the requirement for confidentiality or the payment of a premium fee. Generic hallmarks can also be used to capture new and innovative tax planning arrangements that may be easily replicated and sold to a variety of taxpayers. Specific hallmarks are used to target known vulnerabilities in the tax system and techniques that are commonly used in tax avoidance arrangements such as the use of losses.

Reportable transactions would involve avoidance transactions, which the proposed measure in the Bill would decouple from the definition of “avoidance transaction” found in the context of the general anti avoidance rule. (There would also need to be one of three hallmarks.

Importantly, the reportable transaction rules would be legislated. As is currently the case, they would be in the Act and they would be more general in their description – but they would be legislated and therefore more fixed than notifiable transactions.

The next class is the notifiable transactions. These are much more specific. They target particular areas of concern in tax administration, and they would not be legislated. Of course, there will be legislation going along with them, but particular notifiable transactions would be developed by the Canada Revenue Agency in consultation with the Department of Finance, and so they can be much more specific in terms of the type of transactions that we would like more information on.

Beyond that, the administration and enforcement of the rules are expected to be quite similar, including with quite similar reporting deadlines - but those are the main distinctions.


Tim Barrett – There are two proposed changes on reportable transactions – one is the change in the number of hallmarks required in s. 237.3 – “reportable transaction” from two to one, and the second is that the definition of “avoidance transaction” is going to be effectively broadened to include a “one of the main purposes” test.

Should we expect the existing hallmarks to stay as they are, or should we expect additional hallmarks to be added?

Trevor McGowan – The idea is that the hallmarks in s. 237.3 would be essentially unchanged – confidentiality, premium or contingent fees, and contractual protections.

We appreciate that moving to the one-hallmark test carries a risk of over-disclosure – neither the taxpayer nor the Crown want too much to be disclosed. But by the same token, under-disclosure will frustrate the rules’ policy objective.

When the reportable transaction rules were initially introduced, there was an attempt to try to strike the right balance between those extremes. Unfortunately, our experience has been that the level of disclosure has been, frankly, miniscule, and much below what would be needed for a well-functioning and robust program.

We have heard comments on trying to find the right balance and that is something that we are going to be focused on. Therefore, even though the hallmarks are going to be essentially the same as in 237.3, we will be looking at the specific language to make sure that it continues to be appropriate. In particular, we have heard that the contractual protection hallmark may be a bit overbroad, considering the range of possible contractual protection provisions and insurance that might be purchased in relation to a deal. We are thinking about that sort of thing while developing the draft legislation.

Meaning of "advisor"

Tim Barrett – Who is an “advisor” for the purposes of these rules? Does Finance anticipate creating a due diligence defence?

Trevor McGowan – The measures would build upon existing rules in the Income Tax Act, including the definition of “advisor” currently in subsection 237.3(1), and that section is going to be serving as a precedent for a lot of the new measures being developed or revised. To that end, I would also note that 237.3(11) has a due diligence defence for advisors, so that is definitely something we will be looking at.

Earnings stripping general framework

Phil Halvorson – Can Finance explain the framework of the proposed earnings-stripping rule?

Shawn Porter – We thought it would be useful just to take a few minutes to restate and provide a little colour on the principles that were included in the supplementary information.

First a couple of words on scope and design: the target is generally larger firms – subsidiaries of foreign firms, or Canadian multinationals with significant foreign affiliates. We are contemplating a couple of bright-line exclusions, such as firms with less than $250,000 of interest expenses in their Canadian groups, or for CCPCs with less than $15 million in taxable capital. We are also, as indicated in the supplementary information, not expecting the rule to apply to large stand-alone Canadian firms (i.e., firms that carry on business largely in Canada). The rule would not be expected to apply to stand-alone Canadian firms because the interest expense all fundamentally relate to the Canadian business and its earnings capacity and therefore comes in below the fixed ratio of 30% (or a higher group ratio rule applies). I will say more about the group ratio rule in a moment.

This is not to suggest that larger firms are bad actors and are being singled out – rather, they tend to have more opportunities to situate a disproportionate amount of their global group debt in Canada. That is really the behavioural targeting that underlies this particular measure. This is largely about aligning with what has become the international norm, and recognizing that more debt will be situated in Canada as time passes and more and more countries adopt this kind of interest restriction.

A second point I wanted to touch on is transitional rules. They are intended to provide sufficient time to adjust to the new rules. There are three aspects to those rules: not coming into force until 2023; the relaxed 40% fixed ratio in that transitional year; and the carryback window to pre-regime years. We are aiming to release some draft legislation this summer.

A third area is group treatment and the general framework. As we all know, we do not have consolidated returns or group relief in Canada, so a group relief mechanism will apply as it relates to the application of this earnings-stripping rule. Let’s leave the definition of “group” until the draft legislation is out; we have lots of ground to cover already.

For the purposes of my remarks, I will focus on the default 30% fixed ratio rule. Any reference in that context to “group” and “rules” will be a reference to rules in the Income Tax Act and definitions and provisions in the Income Tax Act, so we are talking about the ITA notions of “group.” We don’t get to other notions of group until we get to the group ratio rule.

The basic framework will be rooted in entities. It will start with an analysis at the entity-level, with the necessary helper rules and infrastructure for the group mechanisms. Each entity will determine whether it has net interest expense in a year. If that is less than 30% of its EBITDA, it has excess or unused capacity; if it is more, it has excess interest which would be denied.

We are contemplating that it is the unused capacity (that is, the attribute that would be transferred to another group entity with excess interest) that allows that other group entity to deduct that excess interest to the extent of the transferred capacity. In other words, the interest is always deducted in the entity that incurred it.

We are also contemplating that all of the unused capacity in a group would be used to the extent of the excess interest in the group in a particular year. Essentially, you take a snapshot of the relevant group, its capacity and interest expenses for the particular year, and combine them within the group.

If excess interest remains in that group in the year after all unused capacity of other group members has been transferred within it, then we turn our attention to carryover rules. The entity with the excess interest can look back three years to see whether there is capacity in the three-year carryback window, and it can also look back three years for other group entities to see whether they had any capacity in that window, to reduce the excess interest in the current year. If there is insufficient capacity looking back, then the denied interest can be carried forward for 20 years and deducted to the extent of the group’s unused capacity in those future years.

There are still design details to be worked out – I don’t want to suggest that this is simple by any stretch. One example is the question of whether unused capacity in prior years is brought forward to a current year with excess interest or excess interest is carried back. Those approaches are substantively similar, but complexities might arise if, for example, one were to carry excess interest back in order to create or increase a loss in that year, with potential knock-on consequences. Another concern, of course, is ease of compliance and administration. Those kinds of details aside, however, that is the general outline of the proposed rules.

Net interest expense

Shawn Porter – “Net interest expense” is broader than the legal meaning of “interest.” It includes interest expense, equivalents to interest expense, and financing expenses. For a general idea of what those “equivalents” will be, I refer you to BEPS Action 4. We’re not doing a carbon-copy of Action 4, but it is good at capturing the various considerations at play.

In general terms, net interest expense is reduced by interest income and equivalents on the income side. Net interest expenses effectively create capacity at 100% of their amount, while ordinary EBITDA creates capacity at the fixed 30% ratio. We are also contemplating that that capacity that emanates from interest income at 100 cents on the dollar would be eligible for transfer to other group entities. Among other consequences, this means that banks and insurers may not have any net interest expense, and may not be subject to the rule.

The Budget supplementary information flags an issue around restrictions with respect to the transfer of unused capacity of businesses of one type to other group members who are carrying on what we might call “unrelated businesses.” We recognize that there will be a line-drawing exercise to help define and distinguish businesses in terms of, say, lending businesses, and imposing restrictions on the transfer of unused capacity that they might naturally have as a byproduct of their business, and restrict their use across the group writ-large including businesses that are not fundamentally lending businesses. We have invited submissions from interested and affected parties on how best to work through those sorts of issues.

Just a couple of quick words on EBITDA: It is generally taxable income. I think this part of it is intuitive, along with adding taxes, depreciation, and net interest expense. If you just think about lost carryovers under section 111, that economic loss in principle has to depress EBITDA at some point over the earning cycle in terms of how the system unfolds, but there may be examples , such as paragraph 110(1)(k), which is an example of a deduction that is fundamentally more akin to taxes than it is to economic loss - so it probably makes sense to have an adjustment to taxable income so as to not artificially depress EBITDA for Division C-type deductions of that nature.

The final point I would make here is around the character of the denied interest as a carryover. The question arises in two contexts: one, when there is excess interest in a particular entity and it is being carried forward (how does that work, mechanically; and what is the character of the interest and how do we achieve the appropriate result, through systematically measuring EBITDA over a period of time?); and second, regarding that excess interest being transferred to another group entity indirectly through amalgamations and wind-ups, so that consideration will need to be given to preserving the appropriate character so that we achieve appropriate outcomes.

Tax loss consolidation

Shawn Porter – Moving then to tax loss consolidation. Typical tax loss consolidation arrangements, whether they are internal debt push-downs, or seeding of deductions in Losscos for the subsequent use or absorption of those Losscos elsewhere in the group, are intended to be unaffected by this rule, which is to say that, subject to satisfying CRA’s administrative practices, their loss consolidation objectives would continue to be achieved. Obviously they get caught up in the application of this rule, but the internal interest expense, for example, needs to be ferreted out of the calculation so as not to muck it up. We are obviously looking at the real external interest expense levels that are incurred or borne by organizations.

The goal is to plan an approach in this regard that is relatively simple and intuitive, but obviously it has to be effective as well. We think there are at least a few ways to doing this. The related party interest income and expense can be made effective just within the Canadian group: we need to find a way to deal with that. As I said, I think there are a few ways to skin that cat, we are just trying to size up which is the easiest to deal with, both in practice, compliance, administration and legislatively.

Group Ratio

Shawn Porter – Not every country has adopted a group ratio. The Canadian rule refers to the ratio of net interest expense to consolidated group EBITDA using GAAP measures for both the numerator, and denominator – but the principle here is that firms can use a higher ratio than the 30% default fixed ratio if that reflects the actual external leverage characteristics of that firm overall.

The idea is that the consolidated financial statements and measures are a proxy for assessing these characteristics. We expect that group ratio to operate well in, e.g., a large domestic stand-alone business or project, such as an infrastructure project that is separately financed and entirely self-contained, as well as in the context of multinationals engaged in integrated businesses.

That is the general idea but, in practice, the application of this rule is going to be complicated. There are challenges and limitations associated with the group ratio. There are a few points I want to flag for you, because this is a good time to be thinking about submissions and engagement on this topic.

One question is how to determine the appropriate ultimate parent – where does the consolidation stop? Who is the ultimate parent in a group context where there may be consolidations going on up a chain? There will be distortions to the extent that the group is loss-making overall, and different kinds of distortions if there are constituent entities within the group that are loss-making. There will also be averaging effects that will be troubling in policy terms in the context of diversified multinationals in very different business-lines with different business and financing arrangements and credit-support. Even more broadly, what if credit-support for the consolidated group is provided by someone outside of the group?

I would refer interested parties to the OECD BEPS Action 4 discussions. We do encourage submissions on this, and we are especially interested in live cases. Situations that you actually think, given the purpose of the group ratio as a proxy for actual external leverage characteristics, that may warrant the benefit of consideration. What is the factual context around a particular situation and why do we have concerns about it, and about the group ratio, in particular, applying?

In summary, I think the principles are relatively clear, but we are aware of how complicated the details can be.

Earnings-stripping rule examples

Example 1a

Phil Halvorson – I have had the benefit of having some discussions where Shawn and Trevor have set me straight on how they expect these rules to behave – subject, of course, to future developments.

The very first example is as very straightforward as we can think of. We are dealing with a 30% group ratio rule. As Shawn said, you want to first determine the limitation or unused capacity situation of each member separately. In this case, Can Holdo has $100 of that interest expense and no EBITDA, so it has not room to deduct any interest, it has a $100 limitation on its own. Can Opco, though, has $800 of EBITDA so, in theory, it could have a maximum interest capacity of $240. It has $200 of its own interest, and so in this case, we are saying the unused capacity is the difference between the theoretical maximum of $240 and $200. Therefore, Can Opco may deduct the entire $200, and will have $40 of unused capacity that can be effectively lent over and used by Can Holdco, so the third step is you start to carry over the attributes for the year and move them around the group, and the idea is Can Holdco could deduct that $40 of otherwise unused capacity and then be left with $60 of limitations that you then consider for the purpose of the carryover rules.

Example 1b

Phil Halvorson – On the next slide, this is the same scenario, except that there is a loan between the two Cancos. Effectively, the $100 of external interest expense of Can Holdco was pushed down to Can Opco. The Supplement states that you ignore the intra-Canada group loans, but only for the purpose of determining the net interest position. When it comes to determining the EBITDA, this intracompany interest is included in EBITDA or deducted from EBITDA as if it were a payment between the two groups that have a non-interest character. So it is a factor in EBITDA.

So what would change between scenario 1a and 1b is not the net interest position, which would remain at $200 and $100 respectively, but the EBITDA of Can Opco would actually be reduced by the $100 payment to Can Holdco, so it would have $700 of EBITDA and Can Holdco would have $100 of EBITDA. With only that change, you get to the same answer: Can Opco would have $700 of EBITDA, and thus a maximum theoretical capacity of $210. It deducts its $200, leaving $10 unused capacity. Can Holdco would have $100 of EBITDA including the intragroup payment, so it could on its own have a $30 capacity. It has $100 of interest expense, so it would otherwise be limited, i.e., $70 of the $100 interest expense would be limited but it can use $10 of the Can Opco unused capacity, so at the end of the day, Can Holdco may deduct $40 of interest, and be left with $60 to carry over.

Note that the numbers in 1a and 1b happened to produce the same result because of the numbers chosen, but that won’t always be the case.

Shawn Porter – I think you have summarized it accurately in terms of what we were envisioning based on the Supps. We do acknowledge that the Supps are a little imprecise. We are still thinking about the best way to do this. As I mentioned a moment ago, I think there are at least a few ways, all of which could be made to work, and the criteria against which we will be evaluating them will be compliance, administration and effectiveness. This is a trickier question than at first appears. It requires stress-testing examples. Loss-consolidation occurs in a variety of different forms. Some scenarios can produce what you might think of as a capacity-deficit in a particular group entity and - left without a forced netting rule - that could give rise to an inappropriate result. These are some of the considerations that we are going through, but Phil, you have captured where we are at, right now.

Example 2

Phil Halvorson – Scenario 2 is very much like scenario 1, except in this case, Can Opco has $1200 of EBITDA, resulting in $360 capacity. The group as a whole has more unused capacity than it otherwise needs, and that result is a $200 deduction for Can Opco, a $100 deduction Can Holdco, and $60 unused capacity that could then be applied through the carry - forward carry back mechanism.

Example 3

Phil Halvorson – This is identical to Example 2, except that we have some unused capacity from the year before.

You must first use all of the unused capacity for the year against the group’s net income for the current year, that would produce a net result. In this case for the year we have $60 of otherwise unused capacity, but we have $80 from a prior year's limitation. The $60 excess capacity accommodates $60 of that $80, with $20 remaining to carry forward to the following year.

Example 4

Phil Halvorson – Can Opco has net interest income rather than expenses. Let’s assume it has no EBITDA. So there is no maximum interest deduction capacity because of EBITDA, but it has $200 net interest income, $100 net interest expense. That $200 net interest income itself is considered unused capacity, leaving Can Opco with $100 of capacity that it can be conferred on Can Holdco.

Consultation process

Trevor McGowan – The dates mentioned below are tentative.

The digital services tax was built upon something from the fall economic statement, it was announced in the budget, fairly detailed budget supplementary information was released and comments were invited on that by June 18. The plan for that is to release draft legislative proposals over the summer, which usually means July but might hit August. It is intended that those proposals would include the core of the taxation rules. These, as it is not an income tax, are currently envisioned to go into a separate Act, and so it is building a new Act from the ground up, similar to what we did with the Greenhouse Gas Pollution Pricing Act a few years ago. After the release in the summer, we would provide feedback, continue to work on those core taxation aspects and in addition develop all of the appeal and objection and background rules that are needed to release a full package by the end of the year, slated to apply at the start of 2022.

In the fall economic statement, the government also announced its intention to engage in consultation on Canada’s anti-avoidance rules, in particular GAAR. That is something we are still working on. We don’t have the specific timeline to announce right now, but it would be great to get it out before the summer when all the other things start to come online.

As for mandatory disclosures, the consultation announced in the budget would run up until the Friday before Labour Day, but the budget did not contain draft legislation or sample notifiable transactions, so these are listed transactions that would be prepared by the CRA in conjunction with Finance. We hope to get them out as soon as possible but certainly with enough time to provide for meaningful conversation and consultation over the summer. One of our main focuses right now is on getting the draft legislation and examples of notifiable transactions out for the consultation.

The interest income-stripping legislation would apply as of July 1st 2022, and we are looking to get draft legislative proposals out this summer. For hybrids, as noted in the budget, we are breaking it up into tranches. There was a commitment to implement the full suite of hybrid measures that have been worked on in conjunction with the BEPS project. Of course, some of those are more relevant than others in the Canadian context. The idea is that we would split it up into two tranches. For the first, we would release draft legislative proposals later in 2021 so that they could apply as of 1 July 2022, and those would be rules implementing the recommendations in Chapters 1 and 2 of the Action 2 Report that generally apply to deduction/non-inclusion mismatches arising from payments in respect of financial instruments, with the second tranche to come out in 2022 and with those rules to apply no earlier than 2023.

As for transfer pricing, that is something the Department has been working on for a while. It’s not just a quick response to Cameco and the idea would be to consult on the transfer pricing rules more broadly. That should be coming in the coming months.