Episode 16: Pareto optimality and its practical application in transfer pricing, with Philippe Penelle

The following transcript has been lightly edited for clarity. Philippe Penelle can be contacted on Philippe.Penelle@kroll.com or via his LinkedIn page.

Intro: Hello and welcome to The LCN Legal Podcast, bringing you expert views and analysis of the legal aspects of transfer pricing compliance. Our focus is always on real-world, practical insights that you can apply in your everyday work. In this episode, we consider a concept that’s well known to economists, but less so outside that field: Pareto optimality. Our guest is Philippe Penelle, a Ph.D. economist with 25 years of transfer pricing experience, including several senior roles. Put simply, Pareto optimality describes a situation in which all the possible value from a transaction has been allocated between the parties, so none of the terms of the transaction can be changed without making one at least party worse off. In discussion with LCN Legal’s co-founder Paul Sutton, Philippe explains how the concept csan be used to create more robust transfer pricing policies for multinational groups, and to counter challenges from tax administrations. We hope you enjoy the discussion – and just to be clear, Philippe’s comments are his own personal views, not necessarily those of his employer or any associated entities.

Paul Sutton: So, hi Philippe, thank you so much for joining this podcast. I’ve been really looking forward to it for a while. So thank you again for sparing the time to discuss this subject with us. And we’re here to talk about Pareto optimal agreements and the related concepts of moral hazard. And I guess more widely, we’re talking about how contractual terms affect economic analysis and vice versa, and how it all needs to fit in with the arm’s length principle. So perhaps you can kick things off just by explaining in brief, what does ‘Pareto optimal’ mean in the context of agreements?

Philippe Penelle: Certainly, and thank you very much for having me. I’m looking forward to this conversation with you. Pareto optimality is ‘economist-speak’ to describe a situation where you have an allocation of resources (in general between multiple parties) and you cannot change that allocation without making at least one of the parties worse off. And what that means is that there’s no value that is left on the table. There is no way – for example, in the context of an intercompany agreement – there is no way that you can either add provisions to the contract or change the provisions that are in the contract without hurting at least one of the parties. Because if you could, it would mean that there’s some value that you could bring to at least one of the parties without hurting any of the other parties, which means there’s value that’s left unallocated between them. And so it generally corresponds to this notion of rationality on the part of each participant to the contract, and the behaviour of profit maximisation – individually, but also collectively.

PS: Right, OK. And this is so interesting, because before I’d read your articles I hadn’t come across this concept at all. So, as a corporate lawyer, I guess it’s a well understood concept in the economics world, but it’s certainly not something that I had come across. So I’d be really interested to hear: when did you first start becoming interested in using the concept of Pareto optimality in the context of transfer pricing?

PP: Yeah, that’s a good question, because it’s really a concept that is intrinsic to the arm’s length principle and the arm’s length standard in the United States. I often compare it with the realistic alternative principle, which is a pricing principle that’s based on rationality and profit maximising behaviour. And the way that you can think about Pareto optimality – which I think lawyers tend to describe as mutually beneficial agreements, and in the kind of economic law literature, that’s the way that it’s referred to – it’s a concept that has always been present. And if you look at the wording, conditions made by or imposed on a controlled taxpayer, that’s really where the concept of Pareto optimality comes into play in transfer pricing.

Because the notion is: uncontrolled taxpayers are assumed to reach Pareto optimal agreements and mutually beneficial agreements. And those are the conditions made by uncontrolled taxpayers. So if you find an agreement and you can show that there’s value left on the table – either you can add a provision that makes at least one person better off and doesn’t hurt any of the other parties, or you can modify a provision that would achieve that. That’s a sign that that contract, by the fact that it wasn’t mutually beneficial, that it’s not consistent with conditions made by uncontrolled taxpayer, which means that these conditions are imposed to the taxpayer by a controlling interest.

And so it is a signal that you violate that condition – that the condition made or imposed by or to the controlled taxpayer must be consistent with the conditions that would have been made by uncontrolled taxpayer. The reality is that before 2010 and before BEPS (2010 is when the OECD published Chapter 9 of the OECD Guidelines), most of the challenges that I saw in transfer pricing were pricing challenges. They weren’t challenges on whether or not the contract was mutually beneficial, whether there was ground for recharacterisation of the transaction. And that picture completely changed after Chapter 9 was published, and pursuant to the BEPS work.

We started seeing more and more recharacterisation challenges from tax administrations. And that’s when I really started thinking hard about the issue of Pareto optimality, because that was a practical way of dealing with those challenges by tax administrations. And those challenges often conflated issues of economic substance that could lead to recharacterisation from other types of, you know – ‘tax administration doesn’t believe that uncontrolled taxpayer would have behaved that way, but cannot show that it is because there’s a violation of economic substance’. So that’s really when I started thinking about it and using the concept to defend clients and to help resolve these controversies that were starting to pop up everywhere.

PS: Right. Got it. OK. And can you talk about a bit about the related concept of moral hazard? And what does moral hazard actually mean? And how does that link in with the concept of Pareto optimality?

PP: Yeah, certainly. Regardless of the property that you’re looking at, regardless of the type of controlled transaction that you’re looking at, when you apply the arm’s length principle, you always have to go through four sequential steps. And you cannot move on to the next step before you have cleared the previous step. And you use the output of the previous step to inform the application of the next step.

And the first step of any application of the arm’s length principle is to remove control and to assume arm’s length dealings. Now, when we talk about control, what we’re really talking about is vertically integrated companies. And there is a reason why companies vertically integrate. The alternative to vertical integration is you’re uncontrolled and you operate through a contract. So it’s a contractual relationship. That’s one way you can do it. The other ways are you create a joint venture or you create a partnership, which are kind of more structural ways of dealing with an uncontrolled counterparty. So when I remove control, what I’m doing is I’m breaking vertical integration and I’m placing my taxpayer in a contractual relationship.

And the question is, OK, you vertically integrated, because it was cheaper to vertically integrate rather than deal through contracting and having to negotiate, potentially renegotiate and all that good stuff. And so when I remove control, when I apply that first step of the arms’ length principle, suddenly I’m creating all kinds of cost that did not exist when the parties were controlled. And I introduce a big issue, which is if you and I are under the control of a controlling interest, the controlling interest will direct us in our actions and we will obey, and the directing party will monitor what we do (and has the ability to monitor us very closely). When I remove control, suddenly everyone is on their own. It becomes very difficult for each of the party to monitor the actions of the other parties.

And that’s what moral hazard is all about. It’s that notion that when you deal with an uncontrolled party, you cannot monitor their actions and they will not make the optimal level of effort unless they are provided the right incentives to do so. I want you to work really hard for me, but I cannot monitor your actions, and therefore I’m going to use the contract to give you an exposed cut of any surplus that a high level of effort would generate, because that’s the way, ex ante, that I can incentivise you to make the optimal level of effort. And I won’t be able to really monitor your effort, but I will be able to monitor what you produce. I will see the sales numbers.

It’s the exact same notion of: if you work as a salesperson, a big chunk of your compensation is going to be directly tied to metrics that are observable, like your sales numbers. And the greater these sales numbers, the higher the compensation that you will get. So when I remove control and I assume arm’s length dealing, I suddenly have to deal with: how am I going to incentivise the counterparty that I can no longer monitor? How am I going to make sure that the counterparty made the optimal level of efforts? Because it’s going to be beneficial for me that the counterparty does so. And dealing with moral hazard is expensive because you do have to negotiate these provisions, and then you have to have a mechanism by which you can verify the metrics that are provided for in the contract, and then you always have the risk that the counterparty at some point will come back and will renegotiate, will force you to renegotiate: ‘I want a higher commission on my sales.’

So all of these costs do not exist when we’re under control. When we remove control, in the first step of the application of the arm’s length principle, we have to deal with these costs, we have to deal with the provisions in the contract that ensure optimal behaviour by each party. And that’s what moral hazard is all about. And it’s a very important aspect to uncontrolled contracting, and therefore it becomes a very relevant aspect of the application of the arm’s length principle.

PS: Right. Got it. My interpretation, in different words, is moral hazard is basically the risk that one party might act in a way which is actually to the detriment of another party: for example, a salesperson who makes no effort to make sales. And having applied the first step in applying the arm’s length standard, which is remove control, it is to create a new structure which addresses that moral hazard and creates a mutually beneficial environment.

PP: That’s exactly right. And another way to think about it is, you know how the language is ‘remove control’ and ‘assume arm’s length dealings’? Removing control and assuming arm’s length dealings put the party in a position where they have diverging economic interest. That is, every dollar that you get is a dollar that I don’t get. So we’re kind of 180 degree in terms of the alignment of our economic interest. You do not want to contract with someone who has completely opposite economic interest than you do.

So you’re going to use the contract in order to start moving the economic interest of one party to get better aligned with the economic interest of the counterparty, because that’s how you maximise the size of the pie, that’s how you get to a mutually beneficial contract. And so the notion of Pareto optimality of a contract is directly tied to how do you deal with moral hazard? How do you deal with providing in the contract the incentives such that the parties no longer have complete divergence of their economic interest, but at least you get them to something inbetween zero degree angle between the two parties, which would mean perfect alignment of economic interest (but you can never achieve that in an uncontrolled contract) and 90 degree, and that’s what you want. You want the contract to take you from 180 degree to 90 degree or less than 90 degree. If you can achieve 45 degrees, you’re great. I mean, that’s a high level of alignment already of economic interest.

PS: OK, so can you give a specific example of a transaction or a situation where this concept can be used to create clarity?

PP: Certainly. And one of the area where that concept is really interesting and important is when we analyse transactions where the taxpayer uses the transactional net margin method, the TNMM, which is often kind of a default method that is used because there’s a lack of transactional comparables available. So we cannot find a specific transaction – and have enough information about a specific transaction – that is comparable to the taxpayer’s controlled transaction. And so instead of using a transactional comparable, we will use a profit-based method, a profit-based comparable, and that’s the TNMM.

And very often, whether the agreement says so or not, the taxpayer is going to implement the TNMM with year-end adjustments, which means that the TNMM would say the range is 2% to 5% for a distributor, for example, and if the distributor ends up at year-end outside of the 2% to 5% operating margin, a year-end adjustment will be made to bring its margin back to 2% if it was on the lower side, or to 5% if it was higher than 5%. When we examine this kind of arrangement, and we ask ourselves what kind of incentive are provided by this type of TNMM mechanism with a year-end adjustment, the analysis will conclude: well, the distributor doesn’t really have an incentive to go above and beyond just what’s required to satisfy the legal obligations that the distributor has, but not necessarily incur the maximum level of effort or make the optimal investments from the point of view of the counterparty who’s leaving the TNM margin.

Specifically, if the distributor works really hard and is particularly efficient and generates, say, a 7% operating margin, the distributor knows that at year-end its profitability will be reduced so that it only ends up with 5%. And on the other side, if you get 1% margin, the distributor knows that at year-end that margin will be increased to 2%. So when we analyse the type of incentive of that type of contracting – and again, even if year-end adjustments are not provided for in the contract; if as a matter of behaviour, that’s the way that the transfer pricing policy is implemented – then basically what we’re saying is that the distributor knows on an ex ante basis that there will be a contingent transfer of value at year-end that will reallocate the taxable income in the transaction between the two parties. And it has an incentive effect, because again, if I know that incurring more effort resulting in higher margin, I will not benefit from it because everything will go back to the counterparty, then I’m probably not going to incur these extra efforts.

So at that point, what you want to think about is the specifics of the transaction. You want to focus on the accurate delineation of the transaction in terms of the commercial objectives of the contract. And here’s where – in transfer pricing we always say that transfer pricing is a matter of facts and circumstances, and we cannot apply transfer pricing as cooking recipes, and this is particularly the case here. There will be many transactions where we will have a commercial objective to the transaction that may be, well, the party that’s contributing unique and valuable assets and function to the transaction – say a manufacturer engaging a distributor in a foreign market to distribute the product – the manufacturer may want, for example, to increase its market share in that market. Or may want to develop from scratch a market share. And these are specific facts and circumstances that you find in the accurate delineation of the transaction that would suggest that incentives are going to be important, and that it may not be Pareto optimal in that specific transaction for the manufacturer to have these year-end adjustments. Maybe a better way of structuring the transaction is actually to provide in the contract that if the margin of the distributor is above 5%, because of extra effort and extra investment made by the distributor, and a distributor that is particularly efficient at generating higher margin, that the distributor will get a cut of that upside as opposed to taking it away from him or her in the form of a year-end adjustment to a TNMM analysis.

And so instead of having TNMM plus year-end adjustment, maybe a superior contract would be TNMM plus leaving a cut of any upside with the distributor at your end. And the same thing on the downside. Because – as much as you want to incentivise the distributor to make extra efforts and make extra investments to be as efficient as possible to meet the commercial objective of the engagement of the distributor by the manufacturer – on the downside, you want to make sure that the distributor will pay the price if it decides to incur a suboptimal level of effort and a suboptimal level of investment.

And so the point I’m trying to make is that none of the statements that I’m making here are statements that you can just apply across all TNMM transactions and conclude that TNMM is never the right method, or that a profit split is always a better method to a TNMM. No, you need to analyse specifically what are the commercial objectives of the relationship between the parties, what are they trying to achieve? And by reference to that accurate delineation of the transaction, you think in terms of how incentives play a role in the structuring of a Pareto optimal contract, or a contract that maximises the value to both parties and is therefore mutually beneficial to both parties.

And in some cases you will conclude a TNMM with year-end adjustment is fine. In other cases you will conclude it may not be the best way of doing things and there may be value left on the table with that arrangement. And instead, you may want to structure the agreement in a different way, which could be TNMM plus a cut of the upside or downside to the distributor, or maybe a completely different mechanism like a profit split, a transactional profit split, which is a more structural way of achieving the same incentive result: both parties have skin in the game and they have better aligned incentives. In the sense that if one party benefit from an upside, it will be shared with the counterparty, and same thing on the downside of the transaction.

PS: Right. That’s really interesting. So, in effect, the concept becomes a framework to apply more disciplined thought to the transaction and to different structuring options. From a contractual and an economic perspective.

PP: I think that’s exactly right. And that’s a great way to put it. It gives you guidance as to what to think about. And I think that element of thinking about incentives, thinking about what kind of incentive a specific contractual provision provides to the party, is an analysis that we tend not to do enough of. But it is particularly relevant when we’re thinking about uncontrolled taxpayers’ contracts and how they would structure the contract and take care of incentive issues, because the contract is the only instrument that they have to do so. And so it is relevant from a transfer pricing perspective. And as you rightfully said, it provides us with an anchor – kind of something that we can hang our hat on to go through that analysis under the specific facts and circumstances that we’ve uncovered through the functional analysis and the accurate delineation of the transaction.

PS: Yeah, really interesting. OK, we didn’t talk about this when we were preparing for this podcast, but one way from a contractual perspective is just to give the principal very tight control rights as regards the activities of the distributor, including the pricing policies adopted by the distributor in terms of sales to customers. What’s your view of that? How would that kind of contractual mechanism fit into that overall concept here?

PP: So that’s a great question, because it goes to the nature of contracting. You only want to put in a contract provisions that would be legally enforceable at a reasonable cost. If the cost of enforcing in court a provision is prohibitive – maybe because it would be very difficult to provide metrics that would give a judge a clear, unambiguous picture, so that the judge can decide the case based on information that you can prove – then you don’t want to put this provision in the contract. And so you have this issue of, if I’m going to provide incentives to another party, for example, I have to make sure that the metrics are objective, verifiable, observable, and that I can gather them at a reasonable cost.

And so the question that you’re asking, and the example that you’re giving: a contract that would provide, say, a licensor control over the price that the licensee sells the product at to the market. We find this provision, for example, in the beverage industry. Syrup manufacturers tend to give exclusivity to bottlers, and by giving exclusivity to the bottlers, they give the bottlers the ability to cut quantity and increase price, which is detrimental. So monopolistic pricing, that’s what exclusivity gives you. You can charge a higher price because you’re protected from competition.

And as a licensor, you definitely do not want a licensee to do that: you don’t want them to cut quantity by raising prices. Because what that does is it makes the licensee better off, because the licensee produces less, so their cost is lower and they can maximise profit that way. But the licensor doesn’t want the licensee to maximise profit, the licensor wants the licensee to maximise revenue, because that’s what we charge royalties on.

And so you see in that industry, these contracts that say that the bottler must fill all orders at the same price and the licensor has a right to cap the price at which the licensee is selling. We have a lot of empirical evidence of uncontrolled contracts that have those provisions. And that goes exactly to the point that you are making, which is, yes, you would use in those cases, the intercompany agreement to give either one party or the other some level of control over the other party. But you are constrained by… I will never be able to really observe how many hours you worked, for example. So that’s not verifiable or that’s very expensive to verify. So we’re not going to structure it like that.

But things like the price at which you sell, the revenue that you generate. Maybe in some cases, the hours that you work, if you must report these hours in a system, like lawyers do. There are certain industries where you may be able to monitor hours, but in many other contexts you can’t. And so you would pick controlling metrics that you can verify and that you can potentially bring to a judge in case of a dispute, and be able to prove your case, because the metrics that are provided for these contingencies are verifiable, and that makes the contract enforceable.

PS: Yeah, got it. OK, so what I’m taking away is, actually, collectively we need to think really quite carefully about control, how that is manifested in contractual terms, and also the structure of remuneration, in terms of thinking about possible incentive type structures. Which is really interesting and something that we touched on last time we spoke, in terms of how profit splits work under the OECD guidelines, and the fact that it’s focused on unique and valuable contributions as opposed to incentives. So could you talk a bit more about that.

PP: Yes. And profit split: you were talking about when did I start really focusing on these issues. And I will say that I spent a lot of time during the BEPS years thinking about profit split. Because in my mind, the BEPS transfer pricing actions, they were by and large driven by the desire of a large number of countries to issue new guidance that would make profit splits a whole lot more prevalent in transfer pricing. And as I was thinking about the arguments that these countries were advancing, I thought ‘they have a really good point, but the arguments that they’re presenting are really not good arguments’. And the arguments in my mind that would have supported their position had to do exactly with what we’re talking about here. Profit split is not just a transfer pricing method in the sense of a valuation tool. It is also a way to structure a transaction and an allocation of risk that you would reflect in an intercompany agreement that would lead you to use the profit split as the most appropriate method.

But you would enshrine in the intercompany agreement that notion that our fate is tied to each other. If there is profit, we will share it. If there is a loss, we will share it too. We are very aligned at that point in terms of the incentives that we have. Whatever is your behaviour, I always have an incentive to work harder, because I will get a cut of any increment. And vice versa: whatever my level of effort, you always have an incentive to work harder. Which means we both have an incentive to work really hard.

And so I started thinking, gee, the reason that we don’t see enough profit split has nothing to do with a lack of fairness (which was one of the arguments that these countries argued for) or with the fact that risk in a value chain cannot be effectively split into chunks that you can allocate. Of course you can do that: you can write a contract, you can allocate risk. So these were weak arguments to support the view that there should be more profit split. But the argument of saying, hey, look, when you remove control and you assume arm’s length dealings, you need to use the contract to better realign the incentives of the parties. And a profit split structure – a structure where you share in the upside, you’re sharing the downside – achieves exactly that. And the reason that we don’t see more profit splits is because people don’t think about moral hazard, they don’t think about incentives when they practise transfer pricing.

And I would go one step further. A lot of people actually do realise all these things. What they do not want to do is have these contingent provisions in contracts, because pricing a contingent provision is much harder than pricing other types of provision. As soon as it becomes a contingent provision, and the value of that provision is derivative, it’s based on some metrics that you identify, you essentially have to use the valuation tool that we use to value options, which is much more complicated.

And so I think on the part of economists in transfer pricing, there’s often a recognition that we shouldn’t ignore incentives, we shouldn’t ignore moral hazard. But the flip side of the coin is if we’re really going to reflect these things in a contract and we’re really going to price them, it’s going to make things a whole lot more complicated. And so let’s avoid those complications, and let’s just simplify and kind of assume away these issues. Which brings you to TNMM. But if you didn’t ignore these issues, you will probably end up in a profit split, because your contract would be structured in such a way that it would naturally lead to a profit split.

Now, I didn’t volunteer this explanation at the time because I was representing my firm at the OECD and US taxpayers at the time didn’t want more profit splits because they had TNMM documentation in place, they don’t want to change it. And the US Treasury at the time was obviously very reluctant to do anything that would lead to more profit splits, because clearly what these countries were trying to do was to get a chunk of the residual income that US companies receive and get a cut of that. And profit split is a way to get a cut of that. They just didn’t go about it the way that could have led to a successful outcome for them.

But I do think – I was very sympathetic with the notion of – yes, the guidance, the transfer pricing guidelines should provide guidance that would naturally lead to more profit splits, but for the right reasons.

PS: Yeah, interesting. OK. And again, just trying to be as practical as possible for listeners. Can you give an example, or maybe a couple of examples, of how you specifically use the concept of Pareto optimality to get a better result for your clients, for the taxpayers?

PP: Certainly. One of the first areas that I use these notions for is in the context of exclusivity versus non-exclusivity of licence. And you know the story. A taxpayer writes a controlled licence as being non-exclusive, prices it as being non-exclusive, but as a matter of fact it is exclusive because the licensor doesn’t license to anyone else. The tax administration comes in and say ‘well you can write whatever you want on the licence’. So they make it an issue of substance over form, or form over substance, whichever perspective you want to take. That’s a very difficult controversy to resolve.

So I started thinking in terms of ‘OK, can I come up with an economic analysis that respects economic substance requirements, Pareto optimality of contract, and prices it respecting the realistic alternative principle in such a way that we could adjudicate this type of dispute with clarity and on a principled basis’. And the answer that I got from that analysis was kind of surprising to me, but very effective. And that was: the tax administration is correct to some degree, in the sense that when a licensee has to make upfront investment to exploit a licence (which in most cases, as a practical matter, they will have to do) of course they want some protection. They want to make sure that they will be able to recover a fair competitive return on the cost that they have to incur.

And so the question in uncontrolled licencing is always a question of: you give exclusivity because you want to incentivise the counterparty to make the optimal investment, and that exclusivity is the way that you protect the return for the licensee. But when you deal with controlled taxpayer, the controlling interest is going to direct the licensee to make the optimal investment. You do not need exclusivity because the controlling interest can direct the parties to do whatever you want. So when you remove control and you assume arm’s length dealing, the now uncontrolled taxpayer finds themselves in the same or comparable economic position to an uncontrolled licensee that would have received exclusivity, because they made the optimal level of investment.

And therefore from that perspective, the tax administration is right. When you look at a controlled transaction and the agreement is drafted as a non-exclusive licence, but it is exclusive in fact (there’s no competition), then the tax administration is right that you should analyse it as an exclusive licence when you get your benchmarks. Because that’s the economic position that reflects the position of the controlled taxpayer, after removing control.

However, then the next step would be: oh, but then of course you can charge more if you’re a licensor in an exclusive licence than if you’re a licensor in a non-exclusive licence, because there’s more value to the licensee and you will extract a portion of that stuff. And the conclusion is: no, that’s the worst way that you can do it, specifically for this moral hazard incentive issue. Because if I increase my royalty – so if, say, a non-exclusive licence is 5% – if I were to raise the royalty rate to 8% in an exclusive licence, I would increase the costs of the counterparty, the licensee. I would give the licensee more incentive to raise price and cut quantity, which I don’t want as a licensor. And on top of that I would make the licence less valuable to the licensee, which means the licensee is much more likely to force me later into renegotiation of the terms. Because if the licence is less valuable to them, it’s less expensive for them to do that. And so the last thing I want to do is increase my cost by providing incentive to the counterparty to misbehave, quote unquote.

So I’m not going to touch the royalty rate. I’m going to keep it at 5%. There should be no difference between the royalty rate for exclusive licences versus non-exclusive licences, everything else equal. But I will use other provisions in the contract to get value back to me if I’m a licensor and I give exclusivity to my counterparty.

And again, the beverage industry is a great industry because we have access to a lot of these uncontrolled agreements and we know exactly how they do it in that case. And there are a number of contractual provisions that you find in these contracts that essentially bring value back to the licensor. But it’s not in the form of cash and a greater royalty. It’s other provisions that give some level of control, as you were mentioning before, by the licensor on the behaviour of the licensee. And that’s the way that that value gets back to the licensor.

If you deal with a counterparty, if you commit to anything, you always expose yourself to being held up later on. It’s true when you’re married, it’s true when you commit to anyone in any way, shape or form, because what that does is you cut off your realistic alternatives and the counterparty knows that, and the counterparty is going to take advantage of that, and you know that when you negotiate. And that’s the reason why you make sure that you put these incentives in the contract to minimise the likelihood that you will be held up later on and all that good stuff. And the way you do that is by not squeezing every penny out of your counterparty, because that’s what makes it cheap for the counterparty to come back and to hold you up and say ‘look, we need to renegotiate and I’m not doing anything. If it takes a year, I’m not going to do anything for a year and you’re going to suffer the commercial consequences’.

PS: So in terms of the outcomes that you’ve been able to achieve, is it really a function of talking the relevant tax authority through the thought process to resist the argument that it’s a de facto exclusive licence and the royalties should be increased?

PP: In those particular case, in my mind, the way that you use this concept is by going through the four steps of the arm’s length principle that we discussed. The first step being removing control and assuming arm’s length dealing, which is when you can introduce this notion that the contract is used to provide incentives. And you cannot just look at the price of the contract. You have to look at other provisions that are incentive-related.

And then you go to step two, which is to test the economic substance. And that’s the place where you have to guide the tax administration to separate out challenges that are really economic substance challenges. Such as, ‘hey, if an uncontrolled taxpayer were to follow this contract, they would breach their fiduciary duty to protect the financial capital of their shareholders’. And that’s a big no-no: lack of financial capacity, not exercising any control, not following the terms of an agreement. All of this would breach fiduciary duty at arm’s length. From these other related notion of economic substance, which have to do with the Pareto optimality of contract, And kind of explain and separate out in the mind of the tax administration the price issue, which is what they’re really focused on, from the value issue.

A contract has value that is above and beyond what’s listed as the price of the transaction. And in the case of the licence, the outcome that we got was there was no adjustment on the royalty rate. Because we were able to show that it shouldn’t matter whether you take exclusive licence or non-exclusive licence. If you had a random sample of the entire population of all exclusive licences and non-exclusive licences, and if you were to run a statistical test of difference on the royalty rates in these two buckets, you will conclude that there is no statistical difference between the two. So: no adjustment on the price, but incorporation – in the other elements of value of the contract – of this provision that you would expect. You keep the royalty rate the same, but you have other elements of value that you need to deal with.

And so we got no adjustment on the price, but we did get agreement by the taxpayer and the tax administration that there would be some changes. I don’t want to call it recharacterisation because it wasn’t used as a punitive adjustment to the taxpayer. But the taxpayer agreed to modify some of the terms of the agreement, make it exclusive – because that’s what it is, as a matter of fact – but also have more clarity in the agreement as to the elements of value and make sure that you had provision that address these elements of value that are not directly tied to price.

So it was more of an issue of clarifying the agreement and making sure that it comported with the whole structure that we’ve just discussed. But there were no adjustments on price. We got rid of adjustment on price that way.

PS: Really interesting, this whole way of looking at things. Even though from my perspective, it’s a very fresh approach, it’s not something that many people write about. From my perspective as a corporate lawyer, it’s directly aligned with the way that we look at the world. Which is, I would say, the golden rule that we apply to any transaction, any intercompany agreement, is: is this an arrangement which the directors of each of the entities respectively can properly approve as being in the interest of that entity and enabling the directors to fulfil their fiduciary duties. So it’s very much aligned with that concept.

We could go on talking about this for quite a long time, but we’re coming to the end. Can I just ask you to maybe offer some key takeaways that heads of tax or tax functions should think about in terms of how they design their transfer pricing policies?

PP: Yes. And I would say, to me, the most important message to companies and tax directors is: the written intercompany agreement is arguably the most important piece of your transfer pricing documentation package. And it is extremely important to have the agreement drafted – not executed, but drafted, and substantially drafted – before you go to step two, step three, step four of the application of the arm’s lengths principle. Step one is remove control, assume arm’s length dealing. What comes out of that should be a well-crafted, clear, unambiguous, written intercompany agreement that can then be tested for economic substance. And if you pass economic substance, the economist takes that intercompany agreement, because that’s what economists are supposed to price when they do transfer pricing. You are pricing the terms and conditions of an agreement that reflects an allocation of risk between the parties. And if you’re a serious economist, you shouldn’t agree to value any controlled transaction without the agreement, because that’s what you’re going to price.

And I think, in our profession, a lot of companies think of written intercompany agreement as being something that you do at the back end, sometimes a year after you’ve started transacting. And they really treat it as a formality, something that you need in your filing cabinets but that is not important. And I think most taxpayers that have gone through real challenges by tax administration – serious audit, and particularly litigation – will realise just how important the written intercompany agreement becomes.

In the US, it’s very commonly the case that judges in tax court refer extensively to what’s written in the agreement, because that’s the only indication to them of expressing what the intents of the parties were. And so, written intercompany agreement: critical. The timing of the agreement: critical. Do it up front, don’t wait until the end. And make sure that it is drafted by both an economist and a lawyer. The lawyers are going to deal with a bunch of terms in the contract that are really important – and that you may not think are important, but have turned out in cases to be important.

For example, choice of law. What law controls the agreement? There have been examples of taxpayers who put choice of law in their agreement – you know, ‘New York law will control the agreement’ – but then they have provisions that provide for a dispute resolution mechanism. And if you actually look at that dispute resolution mechanism, it has its own legal principles that get applied. And so you have conflicting provision.

So you do have terms in an agreement that you should never let an economist ever deal with because they’re incompetent to do that. On the other hand, you have many provisions in agreements that go directly to pricing, that go directly to economic incentives and Pareto optimality, and lawyers are not trained to do that.

And so it’s essential to have a real collaboration between a lawyer and an economist. And they should both participate in the drafting of this agreements. It should be done upfront. And whoever is the economist who participated in drafting the agreement is the economist who should price the contract, because they know what they’re working on. So these are simple steps. It often just means you change a little bit the way that you develop and maintain your transfer pricing documentation. But I think the payoff of making those small changes is very significant. Maybe not at the time you develop your transfer pricing documentation, but as soon as you go to controversy, the contract becomes such an important piece and so determinative in many cases of the outcome.

The better your contract, the clearer it is, the more substance it has, the more you can tie the provision with an explanation as to why at arm’s lengths you would extend those provision, the more likely that you will come out successful from a challenge.

PS: Fantastic. Well, I couldn’t agree more with what you just said, especially as regards the collaborative approach between the economic function, the economists, and the lawyers, which is what we aim to do. So thank you again, Philippe, for spending your time with us, really appreciate it and hopefully we’ll be able to get you back again soon. Thank you.

PP: Thank you very much, Paul. I really enjoyed it. Thank you.

Outro: Thanks for listening to The LCN Legal Podcast. We’d love to know what you think. You’ll find the details on our website: lcnlegal.com. And also on the website, in the blog section, you’ll find a transcript of this episode, including Philippe’s contact details. If you enjoyed this episode, please subscribe. Go to your podcast provider and search for ‘The LCN Legal Podcast’. Until next time, thank you and goodbye.