Brian Ernewein on BEPS
This is an edited transcript of various comments made (in both a morning and afternoon session) on May 26, 2016, at the IFA (Canadian Chapter) Confernce in Montreal, by Brian Ernewein, General Director (Legislation), Tax Policy Branch, Department of Finance. Questions posed by panelists to introduce some of his comments are not reproduced. For the most part, we have not toubled to make editorial changes to emphasize that various references to "you," "we" and "I" were often intended to be hypothetical or non-specific.
Overview of Progress
The title of our panel, The World After BEPS, Phil has reacted to, and I have reacted to as well, in saying that we are not quite done yet. I have felt somewhat contrarian over the last few of years in exhorting people at conferences such as this to pay attention to BEPS (as it might just actually produce something!) - and that people should think about its ramifications and try to provide input on its content - and to try to do the budget consultation exercise as well. And now that we have had these BEPS announcements, and the G20 noticing the OECD last fall, and our budget announcement earlier this year, I have felt the contrarian in saying: Look, those are just the promises – there is a lot of the heavy work to do.
Without trying to take too much time, I would like to run through the list very quickly.
Common reporting standard
The first one I will mention, which is not actually a BEPS-proper measure, is the common reporting standard. That is a big deal. It will be a big deal for financial institutions, in particular, in Canada and elsewhere. We affirmed our intention to proceed with the common reporting standard, in early February at the same time we as signed an agreement with the Swiss that we were going to look at the common reporting standard of that country. That requires draft legislation and then final legislation.
Draft legislation is now out for comment. The final legislation must be be finalized, financial institutions have to gear up to employ this by July 2017, and the Canada Revenue Agency has to set itself up to take that information and to get exchange procedures in place with its treaty partners, so as to exchange the reports that come in under the CRS template.
Country-by-country reporting under the information exchange matter was affirmed in our budget this year, so that we will be proceeding with that. That will require legislation. There is a template in the OECD material, but we are going to need legislation for that. It will affect the largest Canadian multinationals and we will need to get ready for that. This afternoon, I, but mostly Kevin Shoom (who until recently was with the Department of Finance, and is now with the OECD), will talk about the country-by-country reporting.
Transfer pricing/treaty abuse/rulings
On transfer pricing, there are some changes to be reacted to coming out of BEPS on transfer pricing. CRA is applying some of the work that has been produced with the transfer pricing report, and there is more to be done on that score as well. Again, Kevin and I are going to talk about that at greater length this afternoon.
There are changes to tax treaties in relation to treaty abuse and other measures. I will return to that later in this panel. I will spend most of my time talking about the multilateral instrument and the surrounding processes, which are very, very involved.
I know that (coming out of BEPS) CRA is to engage in spontaneous exchanges with other tax administrations of tax rulings that are related to BEPS matters. That has been since the start of this year - so that is something for CRA to take on, and other countries as well.
There also is the BEPS “implementation framework,” which is a sort of peer review exercise whereby we will be assessing other countries, and other countries will be assessing us, on how we are doing in implementing the BEPS matters – or, at least, the minimum standards and other international standards. That is a big deal. At the CFA and the OECD committee on Fiscal Affairs, and its Bureau (of which I am a member), it seems like we have talked about nothing else in the last several months other than bringing more countries in, and bringing in more countries outside the OECD and outside the G20 to join in the BEPS framework. There is a meeting which will be held in late June or early July on this which may be attended by as many as perhaps 100 countries. So there is a tremendous amount of process around that.
You, of course, know about the Panama papers. That has effectively relaunched a lot of discussion around beneficial ownership and I think that will have prospective reverberations.
To return to my initial point, if The World After BEPS title is intended to imply that we are done on BEPS, I beg to differ. I think for us as legislators, CRA as administrators and for you in the tax community and your clients, most of the work probably still needs to be done.
[Several minutes at the beginning of further comments inadvertently not recorded.]
We and other countries are not obligated to adopt the hybrid or interest proposals. That could change, but it is not the case at the moment.
As for the other items (for best practices), I will take the example of the CFC rules.
We thought they were a good sort of prescription for a system that aligns fairly close to our own, as it does with a number of other countries, but there was no agreement on that. We got agreement that people thought, if you chose to go with a CFC regime, the one described would be a good one to have, but that does not mean that there is an obligation to undertake it.
Double non-tax declaration/savings clause
So going very quickly through your list, hybrids are expressed as a common approach but that does not mean there is an obligation to do it. The double non-tax declaration that is to go into our tax treaties, and which will say, in the preamble, that the purpose of tax treaties is to eliminate double taxation and not to promote double non-taxation, is intended to be a minimum standard. [A panelist challenged this statement, but it was subsequently affirmed by Kevin Shoom.]
The savings clause to tax treaties is not a minimum standard. I have my own views on whether we would require it or not, I think Brian Arnold wrote a dozen or 15 years ago and expressed the view that we should have a savings clause in our own treaties, but it is not a minimum standard - nor is the adoption of the anti-PE avoidance rules - and we have already talked about 8 to 10.
Multilateral Instrument process
List of changes
I thought it would be interesting at least, and possibly helpful, to talk about the multi-lateral instrument (MLI) process, which is intended to provide a facility for implementing the Treaty changes coming out of the BEPS exercise.
The entire list of changes is not a minimum standard. For example, we have an aggressive tax transaction reporting regime: there is a question as to whether or not, as a result of the work that was done at OECD, there is something for us to look at and then consider making some changes. That is not an OECD obligation. It is a decision taken by us or other countries by virtue of domestic law in dealing with this list.
We worked very hard on mandatory binding arbitration - and I think the U.S. probably worked hardest of all in trying to make that a minimum standard. There was a real attempt to try to get all countries to commit to mandatory binding arbitration in treaties as a way of resolving treaty disputes. Not that we wanted a lot of arbitrations - but we wanted to have the discipline of an arbitration outlet at the end of the process to force competent authorities to try to get to a decision. That did not happen.
So we are now in a world which is a coalition of the willing. We and other countries are going to be including an arbitration provision in the multi-lateral instrument and pushing for that in our bilateral negotiations - and hope to change the behaviour from the outside.
The multilateral instrument is the proposal that is currently under development to modify 3,000 tax treaties in one fell swoop.
We have (I believe) 15 different treaty changes coming out of the BEPS sessions, a few of which are minimum standards, the rest of which are options. How do you go about implementing those? Bilateral agreement negotiations of all our treaties, would be last down to our kids and our kids’ kids. That is not doable, so the question is whether there is another way it could be done more efficiently.
The objective of the multi-lateral instrument is, very generally, to have a protocol that modifies all of our tax treaties at once, at least to where we would like to have them apply. As I have already said, the scope of this is to cover all of the BEPS-related changes, and so (very, very quickly): there is the transparent entities’ provision for dealing with hybrids; possible changes to the elimination of double taxation provision (especially for those countries which have guaranteed exemptive relief, which can conflict with the hybrid rule - which says you are supposed to be able to tax the residents of those countries if they are not being taxed in the other country or where deductions are allowed – so those exemption rules are to be modified to allow for hybrid application - and also more generally modified to cover off for cases of double non-taxation.)
Outside of the hybrid context, there is a provision to read down or restrict the exemption provision in the standard treaties - and tie-breaker rule for dual resident entities. We have that in our US treaty, and at least a couple of others. A provision along those lines is under discussion to provide two minimum standards (as I recall).
The preamble of the treaty is to be changed to make it clear that its purpose is not just to eliminate double taxation, but also not to foster double non-taxation. Prevention of treaty abuse, the limitation of benefits, or the principal purpose test: this is a big deal, and that is the minimum standard (I will come back to that).
Minimum shareholder rule for shareholder holding periods to qualify for a reduced rate for dividends: we do not have that in any or many of our treaties, but the provision is an option for the multi-lateral instrument.
Changing the rules on capital gains dealing with immovable properties: this would be getting somewhere close to what Canada already has in in regards to treaties.
An anti-abuse rule for third country PEs is the same as the clause I’ve talked about before. Artificial avoidance of PE status for commissionaire arrangements through specific activity exemptions for splitting up of contracts, those are the three things for the PE front (and actually there is more respecting implementation and the like that is contemplated to be covered in this multi-lateral instrument.)
Mutual Agreement Procedure (MAP) - that is a minimum standard for the countries involved in getting their cases of conflict between themselves and their treaty partner to MAP, but does not have as a minimum standard all of the options respecting MAP (e.g., corresponding adjustments, and mandatory binding arbitration) to be covered by the multi-lateral instrument.
Basic approach to choice of options
That is the list. As I said, I think I counted plus or minus 15 different changes. That is a big deal even if you were having negotiations with just one other country. The proposition here is that the MLI is going to deal with thousands of tax treaties to accommodate the modification of any of them in one fell swoop.
What is contemplated to happen under the MLI? The general approach will be that every country that accedes to it is required to adopt the minimum standards.
All the other provisions on this list I have just given, that are not minimum standards, are options. I think that the way that is going to map itself out is that the treaty will cover everything. You will accede to the treaty, and then you will make reservations with respect to any of the optional items which you do not wish to have apply. Now, that seems not too bad so far, but getting into the weeds a little bit, you soon see that there are a lot of questions that come out of this.
For example, how flexible or inflexible do you think it is sensible for this multi-lateral instrument to be? Should you be able, for example, to tailor your accession to individual treaty partners by saying “I’ll take 1 to 6, but not 7 to 10, with respect to countries A, B and C and I’ll make the opposite choice with respect to countries X, Y and Z. If you are doing bilateral negotiations, it will be open to you to do that. You might take the same view in respect of the multi-lateral instrument, but if there are 15 different things, and a dozen of them are optional, unless you just say there is a simple "on" or "off" (yes or no) toggle to the optional items, you will have one heck of a lot of choices, or possible permutations. When you multiply that by 90 or 100 treaties that, say, Canada has, there are a tremendous number of permutations.
The question is whether or not that is all that useful - especially when we have the programming capacity to map all that out for all of the countries. I want to emphasize that discussions are still ongoing, and will be throughout this year. Stephanie Smith is our chief treaty negotiator in the large working group - I think there are 100 plus or minus countries that are involved, which can be a blessing and a curse.
She also is involved with the Bureau, which is involved more with direction setting for that work. I want to emphasize that there are a lot of open questions so far, but as matters stand, there seems to be some buy-in to the idea that we are going to limit choice. We are going to make countries choose their type of platform - minimum standards plus whatever things they want to do - and that is what they have to offer to everyone.
That does not answer everything because maybe there are some treaties that I may not want to do a multi-lateral instrument with (maybe I have some other negotiations, or perhaps there is something unrelated to tax that is informing my views with respect to that country). This represents another question: even if I am obligated to make the same offer to everybody, am I allowed to limit the countries to which I make that offer? In other words, can I exclude them from its application.
Additional issues re choice of options
Other questions are: what happens if the particular treaty already covers the item in question? For example, respecting the hybrid rule, maybe our IV(7) treaty provision already satisfy that. Perhaps we would like to say we do not need to apply the multi-lateral instrument to it and so we would want to reserve on that point. Or maybe our provision goes further than the multi-lateral instrument would. The MLI is not to be a cap on what you can do, and is instead supposed to be a minimum floor. Are you then able to reserve on those particular provisions that are already covered one way or another in your treaties. If you are entitled to reserve, then there are a lot of potential variations.
The question as to whether or not you are allowed to make reservations with respect to your treaties on the basis that you have covered it elsewhere may feed back on the question of whether you have to make the same offer to everybody, whether it might just be simpler, bottom-line, to allow each country to make the offer it wants to each of its treaty partners.
LOB/principal purpose test
On the treaty shopping provision itself, Allison already talked about the choices respecting limitation on benefits: first of all principal purpose test, full-stop; or limitations on benefits with the idea of the conduit rule, or a curtailed simplified LOB. Those are a range of choices. This is a minimum standard, but it is a minimum standard where there is a choice to make. There are a lot of questions even there as to how this is all going to shake out.
For example, let us say that the country Utopia says that it wants to use the principal purpose test. And that is the election it makes, so that is the decision it makes for all of its treaties. And Dystopia, says that it wants to go for a limitation on benefits provision for all of its countries. Thus, there is no meeting of minds, and there is no match - so we that we do not have the multi-lateral instrument operative in that case.
But what if the first country is prepared to say, “I like our principal purpose test best, but I am prepared to take limitation on benefits on the other side.” Will the multi-lateral instrument facilitate that and allow for that degree of choice? What if both countries go into that multi-lateral instrument saying we like our own approach the best, but we are prepared to take the other one. Who gets its choice? What is the procedure for figuring out how to resolve that?
Under so-called asymmetrical application, should it be open for countries to say, I really like my principal purpose test, I really do not want to do anything else, and I know the other likes its limitation on benefits provision, they do not want to do anything else, let us just apply our own test to our own benefits. Is that a possibility or not? I actually think there should be an opening for that. We have had an example of that in the past – Canada did that with the US. There does not seem to be a lot of appetite for it, but I think that it is a question. If you do not resolve that, the question is: do you reject the match and nothing happens.
Challenges of timing
There is a real challenge obviously for drafting the text and over-rides or modifiers given the fact that the MLI will override 3,000 different tax treaties. Each of us is trying to go through our treaties and look at the general language and see if we think it actually does turn it on and whether there should be some sort of schedule or other approach or annex to make clarifications where required in relation to treaties. That remains to be determined but it is a big, big task - and made the more so, I think, by virtue of the process point, which is what I want to close on.
It is intended that this agreement be wrapped up for accession by the end of this year - that is for countries to sign on to it by the end of this year, or the beginning of next. I find that a tremendous challenge to think about. There are some parts of the text where we really have not been involved in their drafting. Just thinking about multilateral treaty and what is involved in that, it seems to me to represent quite a challenge - but that is the game plan of the OECD. I will say this: they have done this in the past and delivered one way or another, so we should not rule out the possibility of that happening.
Limited public consultations
In terms of public consultation, you might be interested that I do not think that there is going to be a release of the text itself for consultation. In terms of the process, and some of the questions which it has raised, the best way to frame this, I think, is that: the OECD is interested in having some form of consultation. There may be something coming out fairly soon on that, which would allow for the possibility of some discussions over the summer.
Kevin and I had already agreed to give him the lion’s share of the time on the panel this afternoon, and I am glad that Kevin joined us and did a review from the OECD’s perspective. I will concentrate on country by country reporting in order to make two or three quick points.
Adherence to OECD framework
First of all, to provide a couple of process points on the design of the Canadian rules to implement country by country reporting, the starting point for our analysis was to adhere very closely to the framework set up by the OECD. I know you are interested in seeing the draft legislation - or should be, to the extent you are part of, or have clients as, large multinationals that will be subject to these rules. The first point is you should expect these rules to be very close to the OECD framework, but you nonetheless will want to look closely to determine if there is anything we should do to deviate in any respect from the OECD model. That would be an uphill proposition from our perspective, but we would like to have that discussion.
Timing of Canadian country-by-country reporting legislation
Next is the timing of this. It will be relevant to you to understand when we are going to get that legislation out. It is up to the government, as it always is, as to how it wants to process budget legislation including in this case the announcement of country by country reporting in the budget. However, it seems to me that the new government is, as with past governments, open to the idea of us proceeding in the way we have in the past. The budget measures did not make their way into the first budget bill - or budget bills in this year’s case. The remainder of the 2016 budget measures (that have not yet been legislated) would, I hope and expect, be released in draft form over the summer as we usually do. I hope that this occurs earlier in the summer, rather than later in the summer - that is, at least, our advice to our government and hopefully they will follow it. So that is what you should expect of the legislation - substantively, and respecting its timing.
Delayed adherence of U.S.
The third point I want to make quickly (although it will take a bit of explanation) is an issue that will affect those of you who are part of, or represent, US multinationals. The issue could apply on a third country basis, but it is really an issue for US multinationals in relation to the country by country reporting. The issue needs background to set it out.
Kevin has already talked about the basic rule: the multinational in its residence (or home) country is generally expected to put together a country-by-country report that goes into its tax administration for dissemination to its treaty partners. However, this requires that country to have the reporting legislation in place (and we have at least one example where that is not the case). If it does not have it in place, local filing could apply, so if you have a multinational with subsidiaries in 50 countries, then each of those countries can demand the report, because the head office company is not being subjected to the requirement to provide the country by country report itself.
To try to manage that, as Kevin just said, there is an opportunity for surrogate filing. The multinational looks around its group, it finds a member of the multinational who is in another country that has country by country reporting in place, and it gets that “surrogate” to file with that country’s tax administration, with that country’s tax administration sharing the report with other countries. That is a way to manage the issue that the home country of the multinational has not put the country reporting in place.
The U.S. has announced that they are going to implement country by country reporting and has indeed issued the Regulations for that purpose (I believe in 2015), but those Regulations are only going to come into effect when finalized and only prospectively for taxation years beginning after the filing requirement - so that there are going to be a substantial number of U.S. multinationals in 2016 that will not be required to file for 2016.
A personal view I will share is that I did not feel that it was the end of the world that 2016 did not get picked up, as we now had the general system in place. However, some other countries felt quite strongly that the commitment was to implement it as of 2016, and felt quite strongly that the US had not satisfied that commitment in all respects by applying it to only some of the 2016 taxation years.
That lead to the concern that the U.S. multinationals might for 2016 be subject to local filing so, to use the 50 country example again, all 50 countries might ask US multinationals to have their local subsidiaries file. As this would be onerous for them, it leads to a question as to whether the US multinationals might try to find a surrogate country to use instead of having the 50 countries all asking for the information.
Surrogate parents for U.S. multinationals
Perhaps they might choose Canada (which is close, friendly, and accommodating). This would be consistent with the country by country reporting regime as the Canadian legislation will stick very closely to the OECD framework. Thus, it would be possible for Canada to be used as a surrogate country for US multinationals for 2016.
This seems "screwy" in that all such US multinationals would come to Canada for one year, thereby giving a work load to CRA which was probably 20 times what they were expecting in relation to country by country reporting.
Thus, there is a question as to what we might do to about that. Nothing has been resolved yet, but one of the things that has been frequently raised is the possibility of the U.S. itself serving as a surrogate country for the U.S. multinationals for 2016 - if the IRS was prepared to accept those country by country returns, and if other countries are prepared to accept that that works in terms of cutting off the local filing obligation if the U.S. uses itself as a surrogate. There is nothing settled on this but work is being done on this, and I am hopeful (given the desirability of a practical approach) that we will get somewhere on this.