Oral argument — Xilinx, Inc. v. CIR

May it please the Court, my name is Arthur Catterall. I'm with the Justice Department, representing the IRS in this case, and I intend to reserve at least five minutes for rebuttal. On one level, this case is about whether the Commissioner's cost-sharing regulations comply with the arms-length standard that governs Section 482. Now, because these cost-sharing regulations implement the statutory commensurate with income standard in a manner expressly intended by Congress, on a more fundamental level, this case is really about how the commensurate with income standard meshes with the arms-length standard. And so I hope to clarify that point today. That would be really well done. Yes. I think most of our treaty partners would. So it seems. So starting with the arms-length standard. You're not from Ireland, are you? No. That's my Irish bro. Starting with the arms-length standard, what the arms-length standard does, it looks at a controlled transaction and asks the question, is this the way this transaction would have turned out? Are these the results that would have occurred if unrelated parties had undertaken the same transaction? That's the general inquiry. Now, given the lack of identical transactions in the marketplace, in the vast majority of cases you look to comparable uncontrolled transactions. And there are pages and pages of IRS regulations. When I say comparability or comparables, it's a highly technical term. Like I said, there's pages and pages of regulations and all sorts of factors you have to look at. So it's not a casual use of the term comparable. By 1986, Congress had become concerned that taxpayers were placing too much reliance on allegedly comparable uncontrolled transactions to produce what they considered or what a court considered to be the arms-length result. And they viewed the problem, they thought the problem was particularly acute in the case of transfers of intangibles, transfers or licenses of intangibles between controlled parties. And so what came out of the 1986 legislation was this commensurate with income requirement, which says in the case of a transfer or license of an intangible, the income from the transfer, i.e., the royalty, must be commensurate with the income that the licensee expects to generate from its exploitation of the intangible. And this was a little bit of a departure from traditional analysis because instead of comparing the controlled transaction with an uncontrolled transaction, now you're comparing the two sides of the controlled transaction. And basically what the commensurate with income standard does, it looks at the controlled transaction. Again, this only applies in the case of transfers of intangibles. It looks at the transaction and looks how the parties have allocated the benefits and burdens of the transaction and asks if uncontrolled parties entered into this same transaction, is this how they would have allocated the benefits and burdens to each party? And the theory being that uncontrolled parties that enter into a license agreement for intangible property are going to strike a deal and just by force of negotiation and economics, for lack of a better word, the benefit and burden ratios of each party are going to tend to be commensurate. Of course, there's all sorts of variables, but that's the way market forces operate. Each side is going to think it's getting a good deal. So when you've got a controlled license arrangement, and let's say, and this is what prompted the commensurate with income legislation, if you've got an intangible that has an unusually high upside potential, it's some very valuable technology or patent or formula or what have you. So if you know that it's got a really out of this world upside potential for the licensee, and if you look to uncontrolled transactions or you try to engage in comparability analysis, you try to find a comparable license arrangement, and if the royalty that you extrapolate from that comparability analysis, if you plug that into the actual deal and the expected or the anticipated benefits in this particular deal, and if you do that and the result is that one of the party's benefits to burdens ratio is disproportionately low and the other party's benefit burden ratio is disproportionately high, then basically what the commensurate with income standard is saying is that those aren't reliable comparables. And so you can't rely on the comparable uncontrolled transaction. You have to, again, look at the economics of the deal and take into account, again, this high upside potential and adjust the royalty accordingly. You know, that's clear hypothetically, but didn't you have a chance in the tax court to show what actually happens between unrelated companies? Well, I think our position is that there really aren't comparable uncontrolled cost-sharing arrangements. Well, then you have to imagine what hypothetical companies will do. Isn't that right? I mean, that's, again, that's sort of what the commensurate with income standard is looking at. But when you get a phrase like what parties at arm's length will do, when you know how they actually behave, why do you imagine a hypothetical? Because it doesn't matter what uncontrolled parties actually do unless that transaction is a reliable comparable to the controlled transaction. Does it say that? I'm sorry? Does it say that? Well, I mean, again, as I said, there's pages and pages of regulations that tell you how to determine, you know, whether an uncontrolled transaction is comparable. You have to make all sorts of adjustments, you know, taking into account risk. Well, don't you have a factual finding by the tax court that unrelated companies wouldn't do that? Yeah. You have a factual. You didn't appeal that. We did not appeal that. And what we conceded was that unrelated parties that enter into, you know, some type of joint development arrangement or joint development activity and that involves sharing of costs generally do not. So why do you think they don't do that? You say it's economically unsound, but why do they behave that way? There could be a number of reasons. I think one theory is that obviously, and the tax court talked about this some, obviously the costs of stock options can be incredibly volatile. There's a great uncertainty about it all. Yes. And the thing is, I mean, I think normally when you've got, you know, two mega companies in the real world entering into some kind of sharing arrangement, that risk, I think they're content to let that risk basically wash. You know, we won't share your, we don't want to take the risk of your stock going through the roof and you having to issue all these options or, you know. Well, you call it a comparability analysis, but it's also what in the real world people at arm's length do with each other. Well, but again, it doesn't matter what uncontrolled parties do at arm's length unless the transaction, I mean, the context they're doing it in is reliably comparable to the controlled transaction. And cost sharing arrangements are a funny breed. I mean, it's difficult to conceive. I mean, obviously there are types of sharing arrangements. You know, unrelated parties, you know, will share, you know, research activities and things of that nature. And basically, you know, whatever they take away from it, it's sort of an intermediate step, and then they go their separate ways and exploit, you know, whatever it is they've gotten out of this. Let me press you a little bit because, after all, the commissioner makes the regulations and says this is generally decided by reference to the results of comparable transactions. Now, if normally a parent and subsidiary, just like the unrelated companies, do it this way, why does the commissioner tell us look at comparable transactions if company, if megacompanies don't do it that way? Well, that's why we're saying you don't look to comparable transactions in the context of cost sharing arrangements. I'm sorry? Why does the commissioner's regulation say you generally do? Because you generally do. In the case of controlled loans, service arrangements, transfers of property, leases, you know, normally in those types of traditional sort of, you know, two-sided transactions, you can find with a, you know, a reliable degree of certainty, you know, comparable, uncontrolled transactions. But megacompanies are different. I'm saying cost sharing arrangements are different. It's, again, it's hard to imagine in the real world, you know, Microsoft sharing all its technology with an Irish startup. You know, here's all our secrets, and we're going to, you know, we're going to share and share alike in whatever comes out of this arrangement. It's just a highly unique type of transaction. They might do it if they had a tax advantage out of it. I'm sorry? If they had a tax advantage of it, even if they were unrelated, they might want to do it. Yeah. I mean, again, I think unrelated parties tend to enter into these, you know, to enter into some types of sharing arrangements, basically to, you know, as a cost-saving measure, to try to, you know, work some symbiotic relationship, efficiencies of economies of scale, things of that nature, more so, you know, on that side, rather than sharing the fruits of whatever comes out of this. Again, you know, unrelated parties, the joint development stage is sort of an intermediate stage. They, you know, they do their research, and then they, you know, they take what they have, and they apply it. They use their own application. I mean, these are competitors, you know. So it's difficult to hypothesize a reliable, comparable cost-sharing arrangement between uncontrolled parties. Is the corporate stock in this case, each of them have separate corporate stock, the Irish subsidiary and the principal company? Do they have separate corporate stock, or is it only corporate stock? They are two separate corporations, yes. And they each issue stock? Well, Xilinx, the parent, is the one that issues the stock options to its employees, and it also issued its, you know, options on its stock to Xilinx Ireland employees. So I think only Xilinx stock is involved here. And is there any comparable circumstance where you would have the arm's-length transaction between non-controlled corporations? Are you talking about just a cost-sharing type? No, no. Oh, oh, oh. Would one uncontrolled party issue options to the employees of the other uncontrolled party? Yes. I'm not a, you know, I'm not an economic expert or business expert, but I'm not aware of. So how are the circumstances comparable? I'm sorry? How are the circumstances comparable? Well, I think we're saying that's one reason that uncontrolled sharing arrangements are not comparable to controlled sharing arrangements. And one thing is, again, in the controlled arrangement when you've got the parent and the sub, only one party can issue stock compensation because, you know, all of Xilinx Ireland's stock is owned by two holding companies, and all of those companies' stock is owned by Xilinx. Xilinx is the only one that can issue any stock compensation, whereas in the real world, I mean, you have two unrelated parties, you know, Microsoft and Dell decide to come together. You know, they're both issuing stock-based compensation. The question is, as far as this intangible cost, is there anything to look to to compare it with? I'm sorry. Are you talking about in terms of measuring it or? I'm sorry. We're trying to find in similar circumstances, comparable circumstances in what you call the real world. Well, and that's what we're saying is that what I think Congress indicated in 1986 and what the IRS has come to conclude is that there are no reliable, comparable, uncontrolled sharing arrangements. And that's why in the qualified cost-sharing rules, I mean, you have a defined transaction, because the view of the IRS is that, you know, you're not going to find, you're not going to be able to produce reliable, comparable, or evidence of comparable uncontrolled sharing arrangements, uncontrolled cost-sharing arrangements. So that's, I mean, that's why you have the defined transaction. And again. So how do you move from there then to, you know, I don't want to push it too fast, but I appreciate the education or the clarification, but you're saying then the substitute for these kinds of transactions, a cost-sharing agreement in a controlled setting, is now you have the consistent with income principle. Yeah. And what the, I mean, when the commensurate with income standard, I mean, it basically operates as a backstop to comparability analysis. So in this situation is applied, you have to then measure the benefits and the burdens. Is that what you're saying? Right. And how does that get established? Well, I think just as a matter of general economic principles, I mean, as a general matter, parties that, I mean, if two parties are going to share costs and benefits, they generally share in proportion. You know, if you're going to share 30 percent of the costs, you're going to get 30 percent of the reasonably anticipated benefits. You know, and I think that's, you know, basically an economic assumption. And I don't think anybody really challenges that. And so, again, when you're under the qualified cost-sharing rules, you've got this defined transaction. And so you may say, well, you know, this is sort of getting away from real world and arm's length and all that. And I think it's important to note that, you know, these rules are not mandatory. You know, Xilinx, if they had wanted to, they could have tried to establish compliance with the arm's length standard outside of the Dash 7 regs. They could have tried to, you know, extrapolate from an economically similar royalty arrangement. And there's, again, pages and pages of rules telling you how you go about doing that. The idea behind Dash 7 is, you know, look, you want to get into a cost-sharing arrangement and you want to avoid the headache of going through this comparability analysis, here's what you do. And it defines the transaction. So it doesn't force it on anybody. And, you know, Xilinx could have, you know, there is no specified method for trying to compare cost-sharing arrangements, controlled cost-sharing arrangements with uncontrolled cost-sharing arrangements. And, again, I think that reflects the IRS's feeling that reliable comparables don't exist. But if Xilinx wanted to, the regulations are flexible enough to allow you to, you know, use an unspecified method. And basically you would just apply the principles, the general comparability principles, which are under Dash 1D. So, again, there are options here. And but Xilinx chose to go under Dash 7. And so it also bears noting that all of this evidence, there's reams and reams of evidence, you know, about what do uncontrolled parties do. My understanding is that none of that evidence was admitted for the purpose of showing that those arrangements were comparable to the arrangement that Xilinx purportedly entered into. They were all that evidence was admitted for the limited purpose of providing foundation for expert testimony as to what uncontrolled parties do, regardless of the comparability, you know, whether it's a comparable situation. So, again, if Xilinx had wanted to, they could have tried to use all this evidence. They could have marshaled all this evidence to try to establish, you know, look, I know IRS, I know you don't think there are any reliable, you know, reliable, comparable uncontrolled sharing arrangements, but we're going to prove you wrong. And, you know, they could have done that. Well, did you have testimony that unrelated companies would do what you imagine them doing? I think, yeah. I mean, I think there is. And, again, I'm not familiar with all the expert testimony. You didn't convince the tax court, but it was there. I'm sorry? You did not convince the tax court, but it was nonetheless in the record. Is that what you're saying? Well, the tax court, I think, confused. Answer that question. Okay. Repeat again. You did not convince the tax court that there was expert opinion that unrelated companies would do what you imagine they do. Is that correct or not correct? We – again, I think you have to look at it in the context of these – when you're looking at what these unrelated parties are doing, are they in similar circuits or comparable? No. Please answer my question. Okay. Let me see some other. Did you convince the tax court that unrelated companies do what you imagine the unrelated companies do? I gather that the tax court – and, again, I think – Can't you answer that yes or no? I gather that we did not, no. No. Now, I want to go on to the treaty, because that's important. Doesn't your position violate the treaty? No. I mean, the arm's length standard in the U.S.-Ireland treaty and the U.S. model treaty is the same as the arm's length standard under U.S. tax law. And, you know, we recently had a model treaty come out, I think, in 2006. And the IRS specifically – or, I'm sorry, the Treasury Department specifically pointed to Article IX, Section 1, which is the general transfer pricing provision and said, you know, this incorporates the general arm's length standard. It incorporates the commensurate with income standard as implemented by Treasury regulations. I mean, so Treasury and – Didn't that say it at the time the treaty was ended? Is that what the law was when the treaty was ended? Yes. No. The treaty was ended some time ago. No. Well, these regulations came out in – I mean, the commensurate with income standard was enacted in 86. The regs, the final cost-sharing regs were 95. The Ireland treaty came out in 97, I believe. And just as a point of clarification, the treaty only applies to one of the years at issue. There's three years at issue. But for that year, aren't you running into the treaty? For that year, we're under the treaty. And the arm's length standard in that treaty is the same as the arm's length standard, you know, in U.S. tax law, in the U.S. model treaty. So the Irish treaty has not been violated. And I'd like to – I mean, the languages between two enterprises that have been those that would have been made between independent enterprises, then that other state. And, again – And as I understand the Irish query about this, we wouldn't do it that way. So – You have to take this into context. I mean, you can't just look at what two uncontrolled parties do. There has to – This is what the treaty says. Well, but I think if you – You can't bring your regulations into the treaty. Well, if you look at the treaty explanation, you know, the Treasury explanation, the Senate report – Do you think we should have the State Department explain the treaty? I'm sorry? Should we have the State Department explain the treaty? I believe it was the Treasury Department that negotiated the treaty. Well, maybe we ought to have the Solicitor General explain it. You know, I think you'll get the same answer. I think everybody in the Treasury Department is on the same page here. It doesn't sound like that. I'd like to reserve a little bit of time for rebuttal unless there are any other questions. Good afternoon. Kenneth Clark, Ronald Squirtenboer, and Mr. Tyler Baker, for exilence. Your Honor, esteemed counsel indicated in his speech just a moment ago that the government has concluded that there are no reliable comparables. Well, it's a good thing that we have judges because the IRS does not rule by fiat. Section 482 cases are fact-dependent. In the lower court, that is the tax court, we presented hosts of comparable transactions. We presented a lot of testimony. The court looked at very specific comparable transactions. It applied very detailed comparability criteria and found that under the circumstances of this case, parties in similar or identical circumstances would not share the cost of ESOs. That's a factual finding. Yes, Your Honor. Let me ask you the same question I asked your opposing counsel. Are there comparables where the stock options are given only to the employees of the one company? Yes, Your Honor. The comparable transactions that we presented, if I recall the evidence, did have stock options where one company was getting the stock option. Yes. Furthermore, we presented evidence regarding other kinds of comparables with government. That's the first example. Do the other employees got no stock options? We had situations in our comparable evidence, Your Honor, where one of the parties to the cost-sharing agreement had stock options. We also presented evidence about private parties contracting with the government in cost-sharing-like arrangements, and in that circumstance, there's only one party that issues stock options. That's the private party. Lots of evidence that government contracts with the private sectors are essentially comparable evidence. We presented an expert by the name of Manos on that subject that is cited in our brief. There was absolutely no refutation. So we presented comparable transactions. We presented evidence concerning the government. We presented expert evidence, and there was nothing in response. So on this appeal, the government concedes that parties would not share ESOs in the circumstances of this case. That should resolve this issue, Your Honor. In the first example you gave, where one of the employees of one of the corporations gave us stock options, the other didn't, did the employees there of both sides get the stock options in the one corporation that issued stock? No, Your Honor. It would only be the one party would issue stock options to its own employees. The question then would be whether the unrelated party would share that as a cost. Yeah. But here, don't all the employees get the stock options? Well, the preponderance of the stock options at issue are issued by and large to the parent company's employees. There are stock options issued to Xilinx Ireland employees, too, but the parent here had far more research and development employees. So the higher proportion, substantially higher, probably two-thirds, would be issued to the parent company employees. So what is that? That just says that they're not trying to allocate all of the costs disproportionate to that, are they? Absolutely, Your Honor. There's no disproportionate allocation of stock options, but if you tried to allocate a cost of stock option, there would be tremendous disproportionate. You would never come close to meeting the commensurate with income standard, which is the predicate. I'm not sure I understand. Explain why that would be the case. Here's how it would be the case, Your Honor. The commensurate with income standard Let's just assume two-thirds, one-third. All right. Commensurate Who of the employees get the stock, how the stock options are allocated across the parent and the subsidiary? All right, Your Honor. Here's how it works. The commensurate with income standard on its face, face of the statute, says the income of Ireland has to be commensurate with the income that the U.S. Xilinx receives. Okay? So if you assume that the payment from Xilinx Ireland to Xilinx U.S. results in income, and you then include stock options, what that means is you'll have tremendous variability in the payment from Xilinx Ireland to Xilinx U.S. And it doesn't matter whether it's two-thirds, one-third, or 50-50, because stock options as the Court found vary tremendously based on exogenous factors, the stock market, what's happening in the world. So you have tremendous variation on Xilinx U.S.'s income, if you assume that that's income. And at the same time, that's in no way commensurate with the flow of income that Xilinx Ireland is receiving. So if you include stock options in the mix, regardless of which company is getting those stock options, you don't come close to meeting the language of the commensurate-with-income statute. Not close. Furthermore, Your Honor, I would point out that the IRS, in its own releases, say very clearly in the White Paper that whatever the commensurate-with-income standard is, it does not impose a formulary result. It does not impose a formulary result. They're trying to substitute an absolute formula here into Section 482. That's not what the commensurate-with-income standard is. In fact, cost-sharing itself, cost-sharing itself is not some abstract, hypothetical concept in Dash 7 that's removed from the real world. Cost-sharing is a very real-world type of transaction. In the White Paper in 1988, the IRS itself said that cost-sharing is used a lot in the real world. The cost-sharing regulations were based off of the real world. That is what the regulations suggest. It's real world. Now, the government now argues that the cost-sharing regulations are some abstract standard, and you have to look at the benefit and burdens. Not true. That's invention. There is nothing in the regulation in Dash 7 that discusses benefit and burden. Before 2002, before the government briefs were filed in this case, there was no pronouncement, no release from the government saying cost-sharing is an abstract model with a benefit-burden analysis. No. Cost-sharing is based on the real world. It always has been. And in the real world, the evidence was overwhelming at the trial court that real-world parties would not share ESOs at arm's length. So what the government is trying to do here is impose an excessive burden on related companies, a burden that unrelated companies would not bear. That's patently unfair. That patently violates the purpose and objective of Section 482, which says the arm's length standard applies in every case, every case. They're trying to create a different standard, which is not an arm's length standard, an evidence preclusion standard, and they're trying to prevent us from looking at similar or comparable arrangements, and that's exactly what the judge did. He refused to be prevented from that, and that is a correct reading of the law. Not only is it a correct reading of the regulations, but it is also a correct reading of the treaty. Kennedy. Could you go back to your two-third, one-third? I'm not sure. I'm trying to track where the in – what's the income attributable to in that? Well, actually, the words of the statute are income attributable. The income attributable to the – to the intangible would be the income received from Xilinx Ireland. Income. That has to be from selling the product. Selling the product. And that has to be commensurate with the income that Xilinx U.S. gets. Yes. And so if you throw ESOs into the equation, and if you consider them part of the income equation, which is highly debatable, but if you do, then the income Xilinx U.S. would receive would go up and down dramatically based on the – That's where I'm having a little trouble. My – I may have slipped – easily slipped a cog on this. My understanding was that the tax court assumed that stock options were a cost. Correct. Okay. So that – that side of the equation I can understand, because the stock has to come from somewhere. You can – you can – you may have a problem deciding at what point in time you value it, but you can presumably arrive at a cost. And if it's two-thirds, one-third, then two-thirds of that cost would be allocated to Xilinx, and one-third to Ireland, correct? Correct. Okay. Now, where do you get into the income side of the equation? If it's income, they're both sharing in the income, are they not? I mean, is Ireland selling separate and apart from the corporation? Well, Ireland's cost-sharing share is based on, in part, its anticipated benefits, that is, the income it's receiving. So where would it be getting income that would be different from the income to the corporation? It would be selling the product that the subsidiary makes, I take it, and then it would receive its revenue stream, and then it would pass on under whatever the relationship between the sub and the parent. Yes, but here's maybe a different way to explain it. Let's say Xilinx Ireland has a steady stream from selling its product. Let's say it's receiving $1,000 per month, okay? If you charge it for ESO cost, ESO cost is tremendously variable. It goes up and down based on the stock market. Why is that the case? The stock is granted at some point in time. It's exercisable, I know, at a later time. Yes, because you're on the discussion about whether you value it as of the grant date or the exercise date. The IRS's original theory in this case was that you value it at exercise. Okay. And the exercise price is extremely variable depending on exogenous factors, not necessarily what the company is doing. So if the income, the cost that Xilinx Ireland has to pay varies up and down based on people's selling price, that has nothing to do with, that's not, that selling price in the market, that variation in the market is not commensurate with the steady stream of income that Xilinx Ireland receives. There's just, there's no linkage. There's a logical disconnect under the words of the statute. You know, how you, how you value it is a separate problem, it seems to me. If you object to the method of valuation, that could be resolved. But I don't really understand the problem of saying that this is a cost. Aren't the salaries of the people who work on these projects, aren't they a cost? Your Honor, I don't think we need to get into the issue of whether it is or is not a cost, because even if you assume that ESOs are a cost, and we presented a lot of evidence that they're not, but even if you assume that they're not, you still, that they are a cost, you still get into this issue. Is Dash 7, the cost-sharing regulation, somehow divorced from the real-world arm's-length standard in Dash 1, and it's not? So we don't really need to trouble ourselves with whether they are or not a cost. The tax court says, we'll assume they're a cost, and then the question is, are the cost-sharing regulations, do they stand out separate and independently from the arm's-length standard, or do they work together with the arm's-length standard? The court said simply, the arm's-length standard applies in every case. Cost-sharing is no different. If we disagreed with that, if we disagreed that the generally doesn't preclude resort to another method of allocation, then does your argument collapse? Do we have to accept that the arm's-length is the only alternative, excuse me, the comparable is the only alternative to an identical? No, Your Honor. We presented evidence other than comparables on the subject. We presented expert testimony. We presented logic. We presented a lot of reasons why parties at arm's length, under similar circumstances, would not share the cost. Even the government's own expert, Dr. Newlin, who helped author the cost-sharing regulations, agreed that if you adopt a formulaic approach, like they're trying to espouse here, you would get a non-arm's-length result. And I want to bring the Court back to the objective, the purpose of 482. It's to get an arm's-length result. It's to put related taxpayers on tax parity with unrelated tax parities. If you adopted a formulaic result based on some notions of the commensurate with the income statute, you would be violating the objective and purpose of Section 482. You would impose be imposing a new indifference charge on related parties that is not imposed on unrelated parties. And I might add, Your Honor, as for the formulaic approach that they espouse, if you look at dash 7, they say, well, this formula gives you the arm's-length result. That's not what dash 7 says. Nowhere in dash 7 does it say this ratio is an automatic arm's-length result. Nowhere. Kennedy. Well, let me go back to what you said about the purpose of the statute. It says that the purpose is for the Secretary to distribute a portion or allocate income so that it reflects the income of any such organization, trades, or businesses. So that sounds like an allocation based on the income of the two corporations. Your Honor, I think if you look at the language of dash 1A, it explains that the purpose and objective of the 482 regulation is to essentially get tax parity. You do that by looking at the arm's-length standard and by having something which generates an arm's-length result. The objective and purpose of 482 would be dramatically violated here if you could separate cost-sharing from real-world evidence. And, in fact, there's no language, again, I would point out in dash 7, that says it automatically gives you an arm's-length result. Well, what do you make of the statute itself, which says that the Secretary can allocate a proportion so as to clearly reflect the income of those organizations? Your Honor, the statute itself is an independent reason that this case should go Zeilings' way, and here's why. The statute permits the authority, only the discretion, to reallocate income or deductions if there is an evasion of taxes or if there's a failure to clearly reflect income. Yes. An evasion of taxes or a lack of clear reflection of income. Right. If parties are operating at arm's length, that is, related parties are doing exactly what unrelated parties would do. There's no evasion of tax. There's no lack of reflection of income. Well, there's certainly a – two corporations working together that are not related corporations are certainly going to each try to get whatever income they can. When you have a superior corporation and an inferior or an owned corporation, they don't have the same interest in maximizing their income. Their interest is in minimizing their taxes. Exactly, Your Honor. That may be true in some circumstances, and that's why the arm's-length standard is built into Section 42. It's inherently essentially a proxy for 42 to prevent that evasion of income if there's a motive for related parties to do that, to prevent that lack of reflection of income, you look to unrelated parties, you look at real-world evidence. The Court looked at real-world evidence. He said, I found essentially, the Court says below, evidence that in comparable circumstances, comparable circumstances, taking into account risks, taking into account the difference of the parties, taking into account contract terms, in those circumstances, parties wouldn't share ESOs, there's no evasion of income. I just don't – I mean, the examples you have – I'm having trouble understanding this, I guess. The example you gave of the government being in a partnership with a corporation, well, obviously, the corporation gives its stock options to its employees, not to the government employees. Correct. So there'd be no reason for the government to share in the cost of the corporation giving its own employees stock options. Here you have a case where the major corporation is giving all the employees stock options. So I don't see how that example is comparable in any way. Here's – here's the comparability, Your Honor. In the government circumstance, with government contracting, the government's regulation here says all costs, according to them, of whatever kind, should be equally borne by the cost-sharing partners. Okay? In the government contracting situation, the private party has wage costs. It may also issue stock options. The government and the private party then co-share the development. That's like cost-sharing, right? Now, logic would have it, if the government's theory were correct, that the government would recognize that you've contributed to that development not only by your cash contributions, but by your employees' stock options. So it would be fair and logical under the government's theory today that the government would say, well, you contributed this amount of cost, hard dollar cost, and purported ESO costs, so that should affect what your share is of the ultimate technology. It should affect what your share of payment is. So it's – it's a very close analog. But what the government says is, you include stock options in your reimbursement request, you go to jail. It's not what's done at arm's length. The government recognizes that. So when it pays the private sector or when it divvies up, allocates the income or the technology from a government project, when it allocates those two, it says basically that the costs of ESOs are not shared at arm's length. They don't even belong in the equation. So there is a very direct analog with government contracting, Your Honor. And if I may, the Court raised the question of the treaty obligations. The U.S.-Ireland treaty explanation is cited in our brief at 64 and 65. Treasury has spoken. We do not need another brief from the Solicitor General here. In the treaty explanation at 64, 65, the government said in explaining the treaty, which was adapted after the cost-sharing arrangement, that the cost-sharing transactions are like any other transaction. They're like any other transaction. Ipso facto, you apply the arm's length standard in the same way. Furthermore, in the treaty explanation, the explanation goes on to say that you look at the specific arrangements like any other transaction. So the U.S. government, before it was a party in this action, before it took a litigating position, explained how the arm's length standard applies to this case. It applies just as it does in every case. You consider the evidence. Now, you heard Mr. Carroll say that Treasury was unanimous on one side. How do you account for them getting into this other position on the treaties? Is it just a different group of people interpreting the treaty, or how would you account for it? Well, I would disagree that the government has now taken a different position on its treaties than it did before the cost-sharing. No, not on the treaties, but his position was they were the commissioner. They agree with the opposition and meant to do that when they commented on the treaty. Now, let's assume you're right. They disagree. How do you account for it? Why would you imagine they would be different groups of people doing it? Yes, I think that the IRS has a strong interest in this case. They see the chance at taxing stock options. It would tax — it would raise — And the people that commented — that did the treaty were a different crowd. Yes. Yes, Your Honor. And not coordinated with the — Correct. And I want to point out something further on the treaties, Your Honor. We put in evidence the treaty that was issued between Ireland and the United States after the commencement with income standard was adopted and after the regulations were adopted. There is also a 1948 treaty that also contains an arm's-length standard that existed before the existing Ireland treaty. So you had an arm's-length standard both before the commencement with income standard and after the commencement with income standard. You're saying that's in the treaty, too? The principle, the arm's-length principle is phrased differently, but it's in the earlier treaty. So both treaties, before after CWI and after CWI, same arm's-length principle. What that tells you is that the commensurate with income statute did not change the arm's-length result. They're trying to get to a different result by applying the commensurate with income standard, and the IRS has said again and again and again that the commensurate with income standard does not and should not give you a different result than the actual arm's-length activities between the parties. That's what they said in the White Paper. That's what they've said in their regulations. So the position here today is contrary to a number of IRS pronouncements. They're trying to use the commensurate with income standard, sub rosa or not sub sobrosa, to essentially overturn the arm's-length standard as it applies to this case. That has not been the government's position. That hasn't been its position in its treaties. It hasn't been its position in its White Paper. It hasn't been its position in its regulation. There's a new regulation, is there not, that's gone into effect that doesn't – I'm not sure if it applies to this case, but has the – has the – Your Honor, there is a new regulation, and that new regulation is extremely illuminating. The new regulation was proposed in 2002. It's quoted at length in the tax court opinion. It became final, essentially, in 2003. If you look at the language of that new regulation, the IRS, by regulation, is seeking to do what it is seeking to do in this case. That's what I thought. So the 2003 regulation does what you say they can't do under the existing. And, Your Honor, that will – And so on a going-forward basis, all that you've been saying would be mooted out in – on a going-forward basis, because now they will be allocating in our two-third, one-third hypothetical. They will, in fact, be doing what you say they can't do if you're – if they're bound by the arm's length. Is that – is that correct? Well, Your Honor, whether the new regulations are or are not valid, I will leave for another day. But if they were valid, they would, in fact, accomplish what they're trying to do here by interpretation. If you're correct, may I – may I make one? If you're correct, they violate the treaty just as much as the government's position now. Am I correct? Pardon me, Your Honor? If you are correct, the treaty will be violated by the new regulations. Absolutely. The treaty trumps the new regulations. Absolutely, Your Honor. And I might further add, if you look at those new regulations, there are so many words in those new regulations that are not in the current regulations. I think a reasonable person would come to the conclusion that the IRS is really trying to rewrite its regulations in this litigation. It took a lot of words to get to the result it's now seeking by its litigating position. So the new regulations are almost in admission by the government in the course of this litigation that its litigation position would not prevail, that it needed to give taxpayers notice, which it never did. Were the regulations subjected to public comment? They were, Your Honor. Did the Irish government protest those regulations? They did, Your Honor. And that protest, or at least they complained about it. And that IRS comment on that is actually in our supplemental excerpts of the records. It's at supplemental excerpt of the records, pages 14 to 16. Were they, was the IRS the only entity for the government that responded to that? For example, did Treasury Department itself or State Department weigh in on that? I don't know the degree of coordination between Treasury and the IRS in issuing the new regulations. Your Honor, usually in the issuance of regulations, there is some coordination. That was my understanding. Excuse me? That was my understanding. Yes. So in any event, several overview points. Yes, because we're coming to the end of the road. We are. We are coming to the end of the road. Commensurate with time. The government's position here, they're really trying to re-litigate the facts, Your Honor. The position they espouse is just not in the plain meaning of Tash 7. You shouldn't cut it loose from the arm's-length standard that has been applied over 70 years in case law. They're trying to apply it much differently here. And all of their arguments about commensurate with income, what it does, what it doesn't do, falls with two basic points. First, the government has said that the commensurate with income standard is not a formula. They're trying to turn it into a formula. Second, the government has said repeatedly that you don't get different results by applying the commensurate with income standard than you do the arm's-length standard. If they applied it the way they would like it to do in this case, if they rewrote their regulations the way they would like to do it in this case, you would get a non-arm's-length result. And that, Your Honor, is patently unfair and violates the regulations as they are written. Thank you, counsel. Just a couple of follow-up points. The idea that the IRS sort of dreamed up this abstract principle in the cost-sharing regs, the cost-sharing regs come directly from the conference report of the 1986 legislation. And I can read it here for you. I mean, the conferees do not intend to preclude the use of certain bona fide research and development cost-sharing arrangements as an appropriate method of allocating income attributable to intangibles, if and to the extent such agreements are consistent with the purposes of this provision that the income allocated among the parties reasonably reflect the actual economic activity undertaken by each. Under such a bona fide cost-sharing arrangement, the cost-sharer would be expected to bear its portion of all research and development costs on successful as well as unsuccessful products within the appropriate product area, et cetera, et cetera. What's your response to the two points that counsel made, one on the two-third, one-third hypothetical and then on the government contract example? Yeah, I'll start with the government contracting example. The context there is, is the government going to reimburse these private contractors for their stock option costs? You know, I would submit that the goal of the government contracting arm is a little bit different than the goal of the IRS, which is to protect the FIS. Why should it be different? If it's fair to charge one, it's fair to charge the other. I mean, it's a cost, and why shouldn't the government pay it in the contracting context? Well, the government has a budget problem. So, I mean, you know, those guys are paid to toe the line. But if it doesn't share all the costs in one situation and wants to charge it, I mean, the government is fairly inconsistent. Well, for one thing, I mean, this isn't, we are sort of talking about apples and oranges here. I mean, these are not, you know, cost-sharing arrangements in the sense of what we're talking about here. Again, essentially, this is the decision, the government policy, you know, are we going to reimburse these people for the, you know, are we going to take this out of our budget? You know, obviously, they've got a different agenda than the taxing authorities. You know, that speaks for itself. And as far as the two-thirds, one-third argument, you know, if the cost goes up, you know, if there's, if the stock spikes up and so there's this huge cost, that's going to affect both parties proportionately. You know, both parties will, you know, will share, will still share costs. You know, their profit margins may go down, but proportionately. There's not going to be any skewing between the related parties and, you know, their sharing of benefits and costs. So I don't quite understand how that undermines the commensurate with income standard. And what about the, on the 2003 regulation, I take it that you would have endorsed the notion that that is essentially doing by regulation what you're arguing is already the principle already? And I mean, that happens all the time. Well, just assuming that, and it was put out for comment with respect to the treaty, what was the inter, what does the record show as to what, I don't know if it's only Ireland, but anybody else, any other tax treaty that may have been affected? Was there a commentary on that? I don't know if the record reflects what, what any other treaty partners might have seen from an unidentified bureaucrat. And, you know, obviously, you know, it's in their best interest to, because, you know, the U.S. is, the income is flowing from the U.S. to Ireland. Ireland, you know, they want the income that they can tax. You know, they've obviously got a different, you know, view of this, take on this than the United States. You know, these things aren't going the other way around. We aren't seeing, you know, big Irish companies set up startups in the U.S. and shift income to the U.S. That's because Ireland. What's, what's the institutional relationship between the IRS and the Treasury Department? We're sort of a controlled subsidiary. Yeah, you could, you could look at it that way. Obviously, you know, I mean, the IRS is within the Treasury Department. There's all sorts of collaboration. On the next door to the DOJ, I was just trying to figure out, you know, they're closer to DOJ than they are to the Treasury buildings. That's, that is true. But, no, there's, there's, I mean, we're talking about a large bureaucracy with many levels of review. I mean, everything is passed by everybody. I mean, that's, that's how the federal government works. I mean, so, again, you don't have the treaty negotiators with one idea and the IRS with another. You just don't. All right. Thank you, counsel. Thank you. Thank you. Interesting argument. Case disargued will be submitted. The court will stand and recess for the day.