At TEI’s 72nd Midyear Conference in Washington, D.C., the Institute conducted a terrific session that focused on the DEMPE functions—development, enhancement, maintenance, protection, and exploitation of intellectual property—as laid out by the Organisation for Economic Co-operation and Development’s (OECD’s) Base Erosion and Profit Shifting (BEPS) Project and its impact on international taxes. Moderated by Cliff Mangano, the expert panel included William Skinner, Michael Lebovitz, and David Fischer. The roundtable below is an edited, condensed account of the session, which, for space considerations, does not include the last twelve minutes of the discussion.
Cliff Mangano: How’s everything today? We have a wonderful presentation. It actually fits the BEPS 1.0 versus the 2.0 you’ve just been hearing about all day. We’re talking about substance [requirements] and DEMPE, and what you need, what you don’t need, and when should you consider it. We have a wonderful panel. I’m going to let them introduce themselves, but closest to me is Will Skinner, Mike Lebovitz, and David Fischer. Why don’t you start, Will?
William Skinner: Hi, everybody. I’m Will Skinner, a partner at Fenwick & West in Silicon Valley. I focus my practice on international planning and corporate transactions, mainly for tech and life sciences companies.
Michael Lebovitz: I’m an international tax partner with Mayer Brown in Los Angeles. Like Will, international tax, M&A, and planning.
David Fischer: Hi, everyone. I’m with Crowell & Moring here in D.C. I’m more of a controversy person, specializing in international transfer pricing, competent authority matters.
Mangano: We’ve divided the presentation into three parts. Part one is substance and DEMPE concepts. One of the things I’ve seen when talking to a lot of people at TEI and other organizations is they are not sure exactly how you define “substance” and then where DEMPE fits in with substance. I think we’ve got on the agenda a way of understanding each concept given the current state of the industry. Then we’re going to talk about their applications to common international structures and, actually, I’m fascinated by this because the three speakers have come up with some examples, some that are being presented by advisors right now. I think it’s worthwhile for everybody to listen to what they’re saying. And we will talk about the proof of substance, which, obviously, at the end of the day as tax professionals, we’re interested in making sure that whatever we do we can prove that it’s correct. We’re going to start with Will talking about substance requirements, an issue with many facets.
Skinner: Thank you, Cliff. So, as Cliff was suggesting, I think we all know substance is important. We know that the era of the Cayman [Islands] entity that owns IP [intellectual property] and has no people and drives profit is gone—long gone. So, we know it’s important, right? And we know we want to have head count; we want to have both quality and quantity of functional support for an international tax structure. But before we get into what that looks like practically, we want to think about it a little more rigorously, because substance comes up with a lot of different facets, a lot of different guises. It’s important to think about, when you’re analyzing this, What am I trying to solve for? What you’re trying to solve for with substance could be a number of different issues. It could be US CFC [controlled foreign corporation] rules, having substantial contribution to manufacturing, or active royalties exception. It could be not having a [permanent establishment]. It could be proving the holding company or intermediate entities eligible for a tax treaty either under a US limitation on benefits or local law. Or, most salient for this talk, it could be transfer pricing and proving out that the IP owner’s entitled to keep its return under the DEMPE concept that came in as part of the OECD’s BEPS Project. Nonetheless, although those rules overlap and the way to solve for them is to have relevant people, relevant decision makers in the jurisdiction, analytically they’re different. We’ll talk about comparing and contrasting that to some extent today.
To set the stage, we want to put on the table an international structure here, which is not uncommon now that the days of that Cayman IP company are gone. We have a cost-sharing arrangement between a US parent and a foreign sub. Most of the people here probably are familiar with what that is, but a cost-sharing agreement would involve a US parent having legal title to all the intangibles, but then splitting beneficial ownership of the intangibles jurisdictionally with its sub. Then, the sub is in a jurisdiction where you can build some substance—it’s in Ireland, it’s in Switzerland, Singapore, pick your country—and it’s a full-blown IP owner and the principal in the structure. OK. Then it’s going to go out and use that IP to generate hopefully positive above-market returns through contracting with affiliates, and it can do that in a number of different intercompany transactions that we show here. Now, in connection with the DEMPE concept, as we’ll see, DEMPE’s relevant obviously if Swiss IP Co.’s licensing IP out to an affiliate, but it’s also relevant to Swiss Co. just acting as a principal or earning income from that IP—you know, making the product, selling the product to customers. As we go through this, this is sort of the structure we’re going to have in mind in thinking about what DEMPE, what substance, does Swiss Co. have to have for your structure to give the results that you’re hoping for. And also, what are the risks you’re trying to protect against? If I lack DEMPE and Swiss Co.’s contracting with the Israeli company to do contract R&D, maybe I’ve got risk that that cost plus return that I’ve benchmarked so carefully doesn’t hold up, and Israel comes in and starts arguing for a profit split. Anyway, we’ll talk about what’s required, but this is sort of the general framework, the type of fact pattern in which you’re going to be dealing with DEMPE.
DEMPE came up as part of BEPS. I can think back fondly—or not so fondly—to when BEPS came out five or six years ago, and all these reports started dropping, some of which became law, some of which did not. And one of the reports that did become law is Action 8–10, which deals with assumptions of contractual risk and ownership of return from intangibles. That became law in the sense that it was embodied in the OECD transfer pricing guidelines, which outside the US, as many of you know, have a lot of credence and a lot of people look to [them] in terms of doing their transfer pricing documentation. The DEMPE concept is in those guidelines, and it sets out certain principles. What it addresses is bearing risk, and it says that if Swiss Co. bears the entrepreneurial risk of developing the intangible, hiring CMs to make it and sell it profitably, it has to have the wherewithal to bear that risk and the functions to control that risk to be respected. Then, specifically to IP, the guidelines say that Swiss Co. needs to perform significant functions related to the development and exploitation of that IP. It’s important to distinguish that DEMPE’s really focused on IP return, although it sometimes gets conflated in practice with other types of activities, such as bearing and earning a return on entrepreneurial risk in general.
So, what is DEMPE? It’s development, enhancement, maintenance, protection, and exploitation of the IP. What you’re going to want to have depends on the business. If it’s a development-stage pharma company, there’s going to be more emphasis on development and maybe protection of the IP. If it’s a marketing intangible, probably more on maintenance and exploitation. What the report says is that if the company’s the legal IP owner, but it doesn’t perform any of these DEMPE functions, it may not be entitled to keep that return on owning the IP. In terms of what you need to have DEMPE, to have the legal IP owner entitled to keep the return, the OECD guidelines don’t provide a lot of guidance. They give some easy examples where the company has no DEMPE, and therefore you lose, but it doesn’t have a lot of granular detail. So, it leaves us as advisors to practically come up with solutions in the business that would work to support a defensible position.
Before we get into that, I’d like to just kind of compare and contrast some of the areas where substance comes up on the US side. And you might be dealing with this in terms of that structure I mentioned, where you’ve got your Swiss or Irish CFC. You want to make sure it doesn’t have a local country issue under DEMPE, but you also want to make sure it doesn’t have Subpart F income, which may implicate certain substance requirements there as well. The US rules are interesting because they actually provide more concrete guidance as far as what’s required, as opposed to the OECD’s more high-level DEMPE concept. As you may know, one of these tests is the substantial contribution test, or subcon test, for an entity that’s manufacturing goods and selling them to a related party. And there, the regs say they want the CFC, through its own people in-country, to perform substantial contribution functions that are listed here. No one function controls, although you could say oversight and direction is a pretty important function, and in the totality it needs to be substantial. For subcon, also the function generally needs to be performed in-country, in the Swiss office, not in a third-country branch, because of the branch rule.
Another area where you can see this that’s specific to IP is the Subpart F active royalties exception, where if CFC’s licensing out IP, or maybe it’s a software company that earns income that’s characterized as software rent for US federal income tax purposes, that’s generally passive income, but it can be kicked out of Subpart F if you establish an active royalty exception. Here, you either need to show regular development and enhancement of the IP—sounds kind of like DEMPE—through the CFC’s own employees, or you need to perform regular and active marketing functions through a substantial foreign organization. On that latter test, there’s a numerical safe harbor. If you’ve dealt with this, you’ve probably run the traps on where you have to show twenty-five percent expenses as opposed to adjusted licensing income. Now, interestingly, on the royalties exception, there’s no branch rule in [IRC Section] 954(c). So here, you don’t need the substance in-country, whereas in DEMPE, you do need the substance, at least in the entity that’s earning the return. Whether you need it in the same country we’ll talk about as we get to it.
The last exception, which probably doesn’t come up as much for you guys as outbound taxpayers, is the active trade or business exception under treaties. Here, there’s a requirement that the entity, if it’s invoking the treaty based on its business connection to the jurisdiction, it has to have active management and operational functions in-country. Again, there’s sort of a numerical safe harbor, so you can look at it and say, “Am I comfortable on a quantitative basis? Do I meet the safe harbor?” And I know from my experience clients that are in here will look at the safe harbor and try to be comfortable that they clear these three ratios. Unfortunately, for DEMPE there’s no such safe harbor. There’s not really even definitive guidelines as to “You need this. If you have this, you’re good. If you don’t have this, you’re bad,” which creates a lot of practical uncertainty for taxpayers in figuring out how to navigate the rules. Like to just put this on the table, because these are rules I think we’ve all dealt with, speaking of my co-panelists, in terms of getting comfortable you don’t have Subpart F, getting comfortable you’ve got a treaty, and you can look to some pretty specific guidance to say yes or no. DEMPE’s a little more Wild West as far as exactly how much you need. You know you need something, but how much do you need?
Lebovitz: Great. So, with that, let’s go back to the structure that Will discussed initially. We’re going to apply these concepts and apply DEMPE to this structure. As you’re building this structure, what often happens is—especially as you’re putting the structure together—these are the questions that we usually get. The first question is everybody’s first question: How many people do we really need to have there? Because this is Switzerland, it’s expensive; it’s Luxembourg, nobody wants to live there; how many people do we really need? We’ll touch on that. The next question is once we got that—three, five, ten, twenty—can somebody commute? Do they really need to live there? Because the person that we really want in that function, our regional lead for this product line, our head of HR, she doesn’t want to live there. She’s happy to commute. She’s happy to go there. You tell me: Is it a couple of days a month? Is it one week a month? What’s the answer?
Fischer: Just to interrupt for a minute, you notice that he didn’t say, “Let’s go to a board meeting once a quarter.”
Fischer: We’re no longer in that world, even though we all remember it fondly.
Lebovitz: Absolutely. Which relates to that—can’t we use service providers for this? You may be using service providers. If you have a Swiss company, you’ll need Swiss resident directors, Lux resident directors. You’ll need actual resident directors, but oftentimes will use a service company for that function. As we’ll see, those people don’t work for DEMPE. The next question is, once we have all that, what level of comfort can you give us? Can you give us an opinion on this? What level of comfort—and we’re going to touch on this in a bit—on proving DEMPE to external auditors, to tax authorities, etc.? And then related to the level of comfort is, can I get a ruling on this? So, with those questions in mind, let’s talk about some common themes. What you should have observed from Will’s comments is some of the overarching themes are substantial management and control. These are real people doing real things, and they have to be capable from an employment standpoint and capable from just a reality standpoint of being able to do those functions. Now, these are all subjective factors. But interestingly, when you go back and you look at active royalties, or you look at the reason why we talked about the active trader business test in a treaty, there are some defined numerical measurements around substantial—7.5 percent or ten percent. There are some defined metrics. They don’t apply to DEMPE, but they are metrics that can be used to provide a proxy for what you might need for substantial. Management and control—again, the people that are in Swiss Co. and Irish Co., they need to be the real deal. They need to be the people that are capable from a corporate authority standpoint, and they have the job description that supports the fact that they are engaged in management and control of those functions. These are what you need for DEMPE—and in fact, whether it’s subcon DEMPE or even active royalties—employees that are capable of performing high-value functions. That’s what DEMPE was meant to be.
Fischer: Can I interject there? When we’re going to look at proving DEMPE later, there are two aspects to this. First, there is who the people are and what their titles are. And remember, some of these people are not going to be available to you later. But, second, we will look at the statistics—how much money are they paid, are they the high-paid people? What is the payroll in-country compared to out-? It’s not just counting heads, it’s counting high-value heads, and one of the ways to measure value is by salary. So, we’re encouraging people to keep track of that, and I know that we all do. But it gets to be a little subjective.
Lebovitz: And whether those functions are duplicated back in the head office. This gets us to the rubber-stamping point: You’ll have somebody in Swiss Co. that is actually sort of making the final decision, but he or she is really just rubber-stamping a decision that has been made somewhere else. The rubber-stamping issue comes up in a number of contexts; it comes up in the treaty context, it comes up in the ECI [effectively connected income] context. But again, these have to be real people doing real things. And there is no number. There is no number. If pressed for a number, I’ve never seen a structure that we or other advisors have gotten comfortable where there’s been less than five people. When you’re building a new structure and you say to the taxpayer, “You need at least five real, serious people,” that’s when the discussion gets serious, because we’ve got to find five real people that are willing to go to Luxembourg and live in Luxembourg and do real things.
Skinner: If I can jump in there, “real people doing real things” is the catchphrase we use, too, for assessing these sort of things. It’s not just quantity but quality, and quality of whether someone is doing real things; you really get a feel for that when you sit down and talk to them for an hour and ask them what they do, what’s your role in the organization. It’s really no substitute when you’re planning for these things to actually spend the time with the people in the business, allow your advisors to spend some time talking with people who you’re going to rely on, because you get a feel for the quality piece. You have to be there with them.
Fischer: Real people doing real things also means that they have to have risk in-country, that they’re paying people. If they have a prearranged stream of income that’s being used to pay them, then sometimes it doesn’t count as being in-country.
Lebovitz: We’ve talked about what are we solving for. As Will mentioned, when you look at that structure, we need to solve for subcon, we need to solve for DEMPE, and if we’re actually paying royalties in, then we have to solve for active royalties as well. And so, these are sort of typical job descriptions, job titles that you’d see to satisfy those particular issues. And you’ll see that there is some overlap, but there isn’t a lot of overlap. What does overlap for all three is management and substantial. But, again, it’s high-level management. You’re solving for different things when you’re solving for subcon, you’re solving for different things with active royalties and DEMPE. There is some overlap, but if there is overlap, it’s with respect to supervision and managerial authority. Solving for subcon, if anybody plays Wordle, that’s playing Wordle. Solving for all three, if anybody’s played Quordle, that’s what this becomes.
So, what are some of the challenges? We’ve hit on this. The employees have to have the authority to do the job that they’re doing. This gets back to the rubber-stamping point. They have to have the authority to be able to do what they’re doing. The problem is that there are different bases on which local countries can attack this. Because, as Cliff mentioned at the outset, DEMPE is IP, and it’s been incorporated in the transfer pricing guidelines in Article 6, which is hard-to-value intangibles. And so DEMPE is intangibles. But, in reality, what countries are using, they’re looking at DEMPE, and DEMPE becomes a proxy for just substance in general. And so, there’s a lot of guidance out there. I mean, there’s I forget how many pages—it’s forty pages in the transfer pricing guidelines on DEMPE. That’s great material for a local tax authority to use even outside the IP context. If you’re trying to prove that Swiss Co. doesn’t have enough substance and so therefore—Will mentioned Israel—Israel says, “Well, some of that value needs to come here through a profit-sharing mechanism.” Then DEMPE becomes a proxy for substance. When you look at the “D”—development—development starts to become tricky when you’re using third-party contract R&D. Who is supervising the contract R&D function, and who has the authority to make the decision with respect to that? A number of companies, particularly in the life science space, as a result of BEAT [the base erosion and anti-abuse tax], restructured their contract R&D, their contract clinical trials, etc., to solve for BEAT. That might create challenges with respect to DEMPE.
Protection, the “P”—oftentimes in a cost-sharing agreement, we’re splitting legal and economic ownership because we don’t want to give Swiss Co. the legal rights because we want to make sure that we can still enforce that IP anywhere we want, and so we want to keep legal in the US. Well, if that’s the case, how do you solve for P? Can you even solve for P in Swiss Co. if Swiss Co. doesn’t own the legal rights that need to be protected? If you split legal and economic ownership, have you lost the P? And then finally, there are issues, depending on your sales structure, on whether you can meet exploitation. So, all that is challenging.
We want to spend a couple of minutes talking about the solution to DEMPE that has been suggested by, sold by, questioned by advisors for the last six, seven years, and it’s basically what’s referred to as the US DEMPE branch structure. Basically, what you do here is [to] see the only thing that we’ve added here to the structure is a US branch, a real branch. This is not a DRE, it’s not a disregarded entity—it’s a real branch in the United States held by the CFC. And so, because it’s a real branch, the employees of that branch are essentially employees of the CFC. These employees just happen to be in the United States doing all of the DEMPE functions that you’d want them to do. They don’t have to move. The only thing we’ve done is we’ve changed their employer. So, they are now employed by Swiss Co. or Irish Co., but they’re still working in the US. The theory goes that, as a result of that, the functions that are performed in the US are imputed to the CFC, and therefore the CFC has DEMPE. And this structure works, provided that you make sure that you don’t have US-source income running through that branch that might trigger ECI, effectively connected income, to the CFC. But this is a structure that has been implemented a number of times, but it’s been suggested as the solution to DEMPE. I will start the discussion off by saying that I’ve been involved in implementing this structure. But that being said, I question whether this structure was intended by DEMPE. Arguably, DEMPE was meant to be taxable DEMPE in the country where the DEMPE functions are meant to take place. The US branch functions won’t be taxed in Switzerland. There won’t be taxable DEMPE in Switzerland. But there are different views on this.
Mangano: Let me ask you a question, Mike: What tax authority are we looking at to prove that it’s DEMPE?
Lebovitz: You’re looking at the foreign countries.
Mangano: Hypothetically, just say you’ve got a Swiss Co. and you’ve got a UK company.
Mangano: Who’s got to agree that it’s DEMPE
Lebovitz: Let’s say you’re dealing with diverted profits tax. Let’s say you’ve got an LRD [limited risk distributor] in the UK. The UK has to get comfortable that Swiss Co. has DEMPE.
Mangano: What we’ve discussed, and your feeling right now is, it’d be easy for the UK to not associate the DEMPE functions in Switzerland.
Lebovitz: Yeah, I mean, I’ve seen this question—and David’s going to talk about this—I’ve seen this question at the external auditor level. And we’re going to touch on that in a second, but that’s where I think initially the challenge has been. So, we’ve seen there’s sort of a split between the accounting firms. Some accounting firms are saying, “Yes, this works on audit,” and some accounting firms are increasingly saying, “No, this doesn’t work on audit, and we’re going to require you to put up some form of a provision.”
A Structural Issue
Fischer: We live right now in a remote world. Looking at the room, there are lots of people also watching and listening to us that are remote. But this is different, right? This is a structural issue. I’m wondering about the line drawing, because I can see that if you’re in Luxembourg, and you have people that happen to be across the border, or even have an office, you have a hub somewhere that has people that are local but they’re not actually working in headquarters, and they’re not actually working in-country, it has a different flavor to me than this branch in the United States—especially if the branch was preexisting, if that was people that were already there that are then being bottled up into a branch.
Lebovitz: This isn’t Belgians going across the border to work in Luxembourg every day, French people going across the border to work in Switzerland every day. This is different.
Skinner: To just kind of make the counterargument on this a bit [laughs], there’s something intrinsically uncomfortable about having a CFC engaged in a US trade or business, even if technically you avoid having ECI. But on this DEMPE issue, if Switzerland has a treaty with the UK, and Article 9 says, you know, arm’s-length pricing, then I guess DEMPE in theory is consistent with that. We could argue that for quite a while separately, but we won’t. If you say that it’s part of the treaty, Article 9 doesn’t have any concept of a triangular case, or the function’s not including branch functions of the related party, from the treaty perspective, that’s more of an issue between Swiss and the US about how much of Swiss’ profit should be attributed to the US branch. I could argue technically I could see why this does ward off DEMPE, although on the other hand DEMPE is such an open-ended thing, you could also see how a tax authority would just find a way to argue it because of the optics.
Mangano: Kind of following that, you got some profit in the US DEMPE branch. UK, in my example, comes to the Swiss company and says, “Well, you don’t have any of the significant people functions. You don’t have any of the control management here, so I’m going to take it and I’m going to tax it in the UK.” If you don’t have the DEMPE branch, how does the US grab that income and say, “No, I have an issue with you [UK], because you’re now double-taxing my company, because you’re taxing the same income I’m taxing.” That’s the part I see in this where the question is. I’m not saying there’s any real answer yet, but where the question is concerned, if you don’t have DEMPE, there’s nothing you could do, right? Because there’s no way to get there.
Skinner: The other advantage of the structure obviously is it’s relatively seamless from a business perspective. Because obviously, in these discussions of “we need five real people doing real things,” there’s a lot of push and pull. Different companies have different priorities. In this case, you’re basically just moving a segment of the US functions under the CFC, and the business isn’t really changing. It’s just the legal entity under which US employees sit.
Fischer: It’s a hole big enough to drive a
Fischer: In the event that this works—and it’s going to get tested of course—it could be applied in many, many areas. We’ve always known that. The DEMPE branch has to have profits, though. So, there is US risk, and there’s US risk that [the] US could start applying BEPS concepts and say those people are here, no matter who they’re working with.
Mangano: Right. So, they’re being taxed in the UK and taxed in the US. Because there’s US profits, you’re paying tax in the US.
Lebovitz: I don’t know. If you’re making sure that that branch only has foreign-source income, then the only way the US taxes that is through subcon or nonactive royalty from a CFC standpoint. But as long as there is only foreign-source income running through it, I don’t know that there is any US income.
Mangano: Depending on what they’re doing.
Fischer: How do you do that, by the way? When you have a person who’s a significant person, somebody that’s running the business, who’s sitting in a branch, and he’s making decisions for Luxembourg, and he’s also making US decisions all at the same time, how do you protect against that being a US source?
Lebovitz: Because let’s assume that the only income going into Swiss Co. are the royalties coming in from the licensees, and they’re foreign-source royalties, as long as you don’t have a US office that is material with respect to producing that royalty income, then that’s not US-source income.
Skinner: Or the other exception you see people rely on is foreign office material participation—
Skinner: —for foreign sales. So maybe the US DEMPE branch is a material factor, but then you check the box on some of the marketing entities, and they show the established foreign office. And those ECI rules, if you plan in them correctly, they’re pretty black and white, or you become very comfortable.
Lebovitz: The challenge here as well is when, as an advisor, do you introduce this idea? We’ve already told the client, “You need five serious people. We’d love more, but you need five people. And we’d really like to see that continue to grow.” Well, then, if you introduce the DEMPE branch as well, then you’ll never have more than five, because you have to have at least five to make sure that you’re solving for the material participation issue. But you’re never going to get more than five.
Mangano: But would you introduce the five, and then the answer would come back, “No, we’re not going to move five.” Then you would introduce the DEMPE.
Mangano: But not at the beginning.
Lebovitz: Would we agree that, even with the DEMPE branch, you still need people in Swiss Co.?
Skinner: Yeah. You need people outside the US, and I think it depends on the facts. I do think the larger point is, though, once you’re in this structure, the tax departments can have a lot of trouble leaning on the people in the business, “Oh, we do really need to move those three senior directors in R&D to Switzerland.” Then you get told, “Well, I thought we had a structure where we don’t have to move those people.” And then once you’re in this, this is sort of what you’re in bed with.
Lebovitz: And some of the interesting implementation challenges with this structure is, now you’re taking a senior person, a senior vice president of R&D, and he or she is moving to Switzerland or Luxembourg or Ireland. And now you’ve said, “By the way, you’re no longer employed by the parent company; you’re now employed by the Swiss company.” So, we’re changing your employment. You’re probably eligible for stock options, so now we’ve got to figure out how to do the stock option plan with Swiss Co. And you say, “Do we really need to do that? Can’t we just make sure he’s employed by Swiss Co., but can’t we just have some kind of service agreement, basically, to put back all of the functions other than his pure employment back into the US?” And now you find yourself explaining to a foreign tax authority why these people are really employees of Swiss Co.
Mangano: Why they meet the DEMPE functions in Switzerland if everything’s somewhere else.
Mangano: I keep hearing “five people,” “top people”—that usually translates to the five people, then ten more people to support those five people, even for just admin, you know? So, the dollar amounts start getting bigger and bigger.
Lebovitz: Right. And that actually is the usual way it works. I had this discussion with a client where it was the exact same thing: “How many people do we need?” And it was the Netherlands. How many people do we really need there? For this particular industry, it was between five and seven. Now they’ve got 100 people in the Netherlands, and that’s the way it’s meant to be. But we find ourselves as advisors trying to sort of sell that structure, just because of all the challenges of the DEMPE branch. I think that’s probably a good segue now to “how do we prove DEMPE?”
Fischer: Sure. So, the first question—and this relates directly to what we’ve been talking about—is, When do we need to prove DEMPE? What I’d like to assert is that this issue is going to exist throughout the life of the taxes, that you need to think about this of course when we’re planning and when you’re dealing with outside advisors, particularly your auditors. But you have to go beyond that, obviously, and think about what might happen if you are examined by tax authorities around the world. So, you’re proving DEMPE throughout this period. The standards are now getting increasingly high for planning, and we’ve seen, and we’ve discussed a little bit, how external auditors will even interview people and will take a very, very hard look at how your structure is working and different attest firms have different views on some of the things we’re talking about. They’re going to take a very hard look, but things get even harder as you go into examination by the tax authorities. Unfortunately, not only does it get a more intense look, but it also becomes more difficult, because you’re now after-the-fact, and people are gone and your documents are a mess. So, as a tax department, we now have to think very, very hard about doing more than preparing simple transfer pricing document packages, and we need to prepare our DEMPE proof so that we can be ready for the future. There’s two themes to this sort of proving side of our discussion here. Number one, be aware when you start that you’re going to end up in front of—or at least assume that you’re going to end up in front of—a tax authority; and second, don’t forget that the tax authorities are going to get information about your structure just automatically through the various reporting country-by-country and so on. They’re going to have some information; be aware of that and take it into account in your planning process.
Lebovitz: And let’s not lose that—advisors give subcon opinions. We give an opinion: “Yes. This CFC meets a substantial contribution test.” It’s less common for advisors to give a DEMPE opinion, but we’ll write a memo and say, “Yes”—we’ll do interviews, etc.—“these functions are consistent with DEMPE.” The question, then, is even if you have that, will your auditors still require a reserve of some fashion? And the diligence point, I think, is especially good, because if you’re making an acquisition and you’re doing the due diligence, the level of comfort you have through an opinion or otherwise is going to be highly relevant in the diligence process.
Fischer: Absolutely. And in opinions that say, “You’ve told us that you have business purpose. Thanks.”
Fischer: Maybe more than ten years ago, I suppose, we could live with that.
Fischer: That doesn’t get very far in front of the tax authorities, and it doesn’t get very far anymore in front of your auditor, either. Looking to what’s happening down the road, once you’ve done the planning, you have to think about where enforcement is going to happen. Because we’re US-centric, we always start with the United States, but we’re going to go over that very quickly. DEMPE is not a US concept right now—substance is a US concept, but DEMPE has not been adopted in the same way that it’s being adopted around the world. I’m going to mention only in passing, because I’m sure all of you have heard people speak about Coca-Cola [and Subsidiaries v. Commissioner (2020)] many, many times, but the Coca-Cola case shows how the US tax authorities are successfully arguing substance these days. In Coca-Cola, they said payment for activity without actual control or the ability to control activity is not enough. Assigning risk after the fact, once you’ve already got an income stream, and then saying, “OK, now it’s their risk,” is not real risk. It’s got to be an entrepreneurial risk that the contractual relationships, if they have a terminate-at-will provision, are scrutinized on a different level. The focus through the Coca-Cola case and other transfer-pricing cases that we know remains on substance, and it remains on substance in the IP area even though DEMPE is not officially the policy here yet. Looking to the future, five years from now, things are going to be different. That’s what we’re planning for.
Skinner: I would just briefly echo that, David, that we’ve seen DEMPE questions creeping into audits and IDRs, with asking about people functions and so forth, even where it’s not technically legally relevant under the 42 Regs, because we have this rule that generally the transaction that you set up has to be priced unless it lacks economic substance, which is, in my view, a much higher bar than for someone to disregard something in DEMPE. But nonetheless, we do see the examiners starting to raise these type of questions in US audits.
Fischer: And it comes in because of the company’s own documentation, too, because everybody’s dealing with countries on both sides of that coin. And when you’re dealing with DEMPE elsewhere, it naturally finds its way into your discussion, so you’re raising it as well. Non-US DEMPE enforcement is interesting, I think. It comes in sort of different flavors, different regimes, and I kind of labeled them here, just for purposes of discussion: reporting, targeted taxes and anti-abuse, but these all sort of blend together, and I’m doing this for convenience for talking. Country-by-country is an obvious example of reporting, but I’m going to skip over that for a minute and talk a little bit about the practical compliance guidelines in Australia. The practical compliance guidelines have existed for a while in Australia in transfer pricing and other international areas. I think the first ones were with respect to debts and so on. But if you are required to file a reportable tax position schedule in Australia—meaning you’re a certain size, which in our world is not actually that big—then you may be required to report on your risks. They have a high, medium and low risk scale. The newest practice compliance guideline has to do with this very thing—intangibles, intangibles arrangements—and the practice guideline is worth reading. If you haven’t read it, it’s a little bit long, like everything in this area—there’s always fifty pages on DEMPE everywhere—but it goes through the things that I’m more worried about, which is how do you prove it? What kind of documentation should you have? What kind of data analysis should you have? It walks through a lot of that in line-by-line detail. The practice compliance guidelines say to taxpayers, “Rate yourself high risk, medium risk or low risk.” You’re doing your own audit choice, right? So, of course, we’re all low risk, right?
Mangano: But the interesting thing here is it gives the ATO—Australian Tax Office—a starting point, that they’re telling you, “This is how we’re going to view it.” Now, it may be real, it may not be real, but they’re giving you a point, and they don’t tell you how they got to their solution. It’s just how we’re going to view it. Country-by-country also gives them a starting point for “Who do we go for first?”
Mangano: It’s a little different than what we’re used to here, which is, OK, here are the rules, and we don’t really know where they’re going to come unless we are losing money on the transfer pricing situation. And the reason why I think that’s important, though, is Australia seems to be leading a lot of the foreign offices. They’ll put something in place and otherwise adopt a variation of it, and I find it fascinating, because it’s a country of twenty-five million people. I’m from Texas; we have twenty-nine million. And you say, OK, they’re leading in this area.
Fischer: They are.
Mangano: Be aware—if they do it, somebody else is going to do it, particularly in the commonwealth.
Fischer: And this goes beyond, because they look to things like are you forming a new structure? Can you identify a commercial reason? Is it rational, as you said earlier, for this? They list a lot of DEMPE factors and go through whether you’re following the DEMPE factors and how big a tax impact it is. And that’s part of their structure, for you to first rank yourself high, medium, low, and then they come in and you’re supposed to keep the information that’s on 70 pages worth of stuff, to be able to document if they come and ask you about it. Australia is a very interesting example of maybe the sign of things to come in the enforcement area. On the targeted taxes side, you’ve already heard the diverted profits tax mentioned here. I call this a targeted tax because it’s a special tax that’s actually not an income tax. Is just a toll that applies if the transfer pricing authorities find that your international transactions lack substance, have DEMPE issues. They can impose a 25% tax on the profits that have been artificially diverted from the UK, and that rate is going up over time as the rates go on. So, if a UK tax resident company enters into a transaction lacking economic substance, the UK can impose an extra tax, and there’s lots of rules involved in that. Of course, it has its own elements like the BEAT tax, right? That is, if you have outbound transactions that are base erosion-type payments, then you could be subject to the diverted profits tax. Again, I think Australia has this as well; they’re a little bit ahead of what we’re doing here, and they’re ahead of even the OECD. This is something that’s been around for more than five years.
Lebovitz: Yeah, see, I would look at the diverted profits tax as sort of a super-DEMPE. It’s the way the U.K. essentially has legislated in DEMPE. The hope is that the diverted profits tax is one of the taxes that gets eliminated as a result of Pillar One. I’m skeptical about that, but, in any event, the DPT audits that we’ve been involved in, it’s basically DEMPE.
Fischer: Right. And in the distinction between that and an anti-abuse rule, in my view, the general anti-abuse and -avoidance rules require the tax authorities to find a way to tax you. They allow some pretty aggressive changes in your form of your structure to try to follow substance. Myself I’ve been involved in sort of the 245 Canada audits, which is a version of the GAAR, and it’s not as easy for the tax authorities as just a targeted tax, because they do have to find a reason, but it’s somewhat the same in the sense that—it’s something that backs up a transfer pricing audit. If they can’t find a way, right, because they’re unhappy that your branch is in the United States and is giving something substance, they’ll find a way, right? These are the “find a way” rules that are existing around the world. So, we have to be worried about those upfront, and, of course, they’re going to keep controversy people busy for years to come. Country-by-country reporting I’m going to sort of skip over. We all know this. But they’re getting data, right? And the PCG in Australia collects a lot of data, too. There’s a lot of places that are doing this. But some of this is just the numbers of employees and profits and taxes, so it helps identify targets for a country. We have to remember that they’re all getting this, and we’re exposed as a result of it.
Lebovitz: People tend to forget that C-by-C was just meant to be a risk assessment tool, but it’s becoming increasingly the first tool used by foreign tax authorities. We’re all aware of the deficiencies of C-by-C. You can sort of get why C-by-C came about. To the extent that anything in BEPS made sense, there was some logic to this, but the way the C-by-C rules work is they don’t really provide the right kind of information because of the way the information is aggregated, how it consolidates, there are no eliminations. So, the challenge here is that we’ve got a tool that doesn’t work all that well, and yet you see around the world, particularly in the EU, to double down on this and to make this public. And that’s just going to increase the challenges around C-by-C reporting.
Fischer: Absolutely. When this first came out, we were all sitting on panels like this talking about the horrors of apportionment. And now we have the horrors of apportionment coming at us directly through the push.
Fischer: It’s a new time.
Mangano: If you go back to the C-by-C slide, the one bit of information in where the branch structure where we’re talking about is number of employees.
Mangano: And that’s going to be still one or five, in let’s say Switzerland, will maybe have twenty people, IP developers in the branch.
Lebovitz: Correct. Exactly. And it’s basically item seven there that led to one of the impetuses for DEMPE branch, because that’s the first thing somebody’s going to look at because that’s not readily apparent.
Skinner: The other thing maybe, to chime in here, is in planning for substance. You know, really what’s probably legally most relevant are those high-quality individuals who are exercising control of the relevant functions. On the panel, we were talking, we’ve all seen people where you’re adding some number of employees because it’s just better optics, even if they’re people who are performing not high-value functions. To have twenty people who do QC or customer service or whatever, it’s a helpful fact, you know.
Mangano: Or a little different, even if you put the right decision-maker there, but his development staff and software development’s outside the country, you’re going to get into the arguments, “Yeah, but how much does he really do right there?” Head developer in India is doing the work for you.
Lebovitz: If you want to live in Switzerland being the vice president, senior vice president of DEMPE, I mean, that could be a fun role [laughter].
Mangano: Until winter hits.
Fischer: We’ve even seen employees that are not even involved in the same functions at all.
Fischer: They’re in a different line of business, and they’re in-country to try to get them onto the forms.
Mangano: We have five minutes—I just thought I’d give you a warning.
Fischer: What needs to be proven, we’ve already covered this in pretty great detail already. Michael covered it on a previous slide. Obviously, we’re focusing on sort of development protection and exploitation. It tends to be that development is going to be your most important element here when we’re thinking about IP transfers. A lot of that is designing control of R&D functions. We’ve seen that that’s very, very difficult to establish once there is a R&D function in another country, trying to move it through moving a few people, even if they’re the people in charge, it gets to be very, very difficult to prove that that’s actually being controlled elsewhere. But control over budgets and control over strategic decisions is what it’s all about. We talked earlier, it depends upon the nature of intangible, it depends on what kind of business you’re talking about. The decisions regarding enforcement, I didn’t 100% agree with what was said, that everybody wanted to be in the United States, because sometimes they don’t. But the cost-sharing arrangement has sort of shifted that around a bit. And everybody is sort of doing that, from US to foreign, and then we’re involved in hubs outside the United States.
Lebovitz: There’s an interesting overlay between IP rights and legal protection and where you need to have legal ownership in order to be able to enforce, and it depends on whether you’re dealing with a patent or a trademark. And there’s different enforceability, standing to sue, etc., there’s different enforceability issues.
Fischer: We’ve been involved with drafting those licenses that are sufficient to file suit and don’t do much for our purposes for transfer pricing. Sort of the last point here—and this is where I started in my little section of this talk—is that when you’re planning, you have to be looking to the future. Transfer pricing has always been about facts, and DEMPE and substance just makes it absolutely more so. I would encourage people—you know, we go beyond what you need for a document package in the United States, penalty protection. We think that your corporate entity documentation has to be pristine, your transactions and your contracts have to be reviewed for this tax substance. You should have your people keep minutes of even R&D meetings and so on. We find that it’s very difficult to find real proof of what was going on in R&D in the past, other than to the extent that people have R&D credits. And then it’s always in the United States, so that’s a little difficult. You want to keep your correspondence to the extent you can about key decisions. We’ve had companies who, when you have multiple subsidiaries, keep a memo of that subsidiary’s activities for the year in a file with its financial statements in order to know that the tax lawyers can come along five years later and say, “This was what was important.”
Lebovitz: That’s a good idea.
Fischer: It’s been very, very effective, and it was just chance. You know, we didn’t plan this ahead of time, but we had it, and it was a great chance. You need to keep your data, of course. You need to keep track of who are the witnesses, who could be witnesses, who are the younger people that are in the R&D department who could end up being witnesses five years from now, because there’s going to be a controversy later on. Fortunately, with witnesses, in this area, the locals get to do what they’ve always wanted to do. We’ve always been in transfer pricing cases where the witnesses are building up how important they are when we’re telling them, “No, no; you’re not doing anything.” Right? With respect to DEMPE, we get them to say how important they are as long as they’re in-country, right?
Mangano: David, you said that it was by chance you got those records. I would put it on good planning by internal tax department [laughter].
Fischer: Yes, of course [laughter]. That’s what I meant to say.
Mangano: We’re running out of time, but we want to open up the floor for questions because you guys are here—they’ll be around afterwards. There was a couple of questions. I’ll stick to one question. It’s a little long, but somebody asked the question online: How much of an issue is there if the US is ultimate decision-maker and cost-share participant is a Swiss Co., while Swiss Co. is the most substantial entity outside the US? In other words, can US invoke DEMPE concepts when following the cost-sharing rules? I think to summarize, how does cost-sharing affect?
Lebovitz: I guess that would be how relevant is DEMPE?
Mangano: If you look at our example, our org chart, cost sharing’s between Swiss Co. and the US.
Lebovitz: Right. So, how relevant is DEMPE in, I guess, the buy-in, the PC—I just dated myself—it would be PCT.
Mangano: Or the bottom?
Lebovitz: Yeah, but I think the question probably gets more to how relevant is DEMPE to testing the buy-in. That’s a good question.
Skinner: I mean, there are the very prescriptive rules on how you value a PCT, right? Ideally, you’d have some substance under Switzerland to help combat the IRS in the audit. You know, the good thing is for US purposes is you can check the box to get substance. You don’t actually need Swiss Co.
Mangano: And the interesting thing is we’re looking at it from a US perspective, and we go back to DEMPE’s not a US concept.
Fischer: Cost sharing is actually anti-DEMPE in a sense, because what it’s doing is it’s saying you can split the costs wherever they’re being performed, the R&D work, you can split it among the participants.
Lebovitz: But it still begs the question of who’s incurring those costs and where are they being incurred? So, you can say, OK, because of these functions, the IDCs are really in the US.
Mangano: To use an old term, we have a cliffhanger, right?
Lebovitz: Oh, there we go [laughter].
Mangano: If you have any questions, they’ll be around afterwards. I’d like to try to stick to the time schedule as well as possible. Thank you very much to our speakers. And they’ll be here afterwards.