Skip to main content

The Story of Cost Sharing Agreements

Mimi Song: Let’s start with the basics. What is the cost sharing arrangement and what is the purpose of using one?

William Byrnes: For those of us who are lawyers in the room: words have meaning, the tax code has meaning, and we’re not China. We’re not Venezuela. It doesn’t just happen on the fly, in a discussion with government. We have a book and we rely on that book to make business decisions. Tax is one aspect, but just one.

So by example, we could go through the formal definitions of cost sharing, but I broke down just this basic [view]. I took all the different regulations. It’s it comes down to it’s an aspect of negotiating, a joint venture between at least two parties.

Matthew DeMello: And just to interrupt quickly, Fiona, how have cost sharing regulations changed over the years?

Fiona: The definition of intangible broadened to include workforce in place, goodwill, and going concern value. It also permitted the IRS to value intangibles on an aggregate basis in cases where it achieved, and I quote “a more reliable result.”

So you can see they kept the definition pretty loose on purpose.

William Byrnes: That’s the most basic cost sharing [definition].

And then it has three elements 1. For the mutual development of intellectual property. 2. The protection of IP rights potentially contributed. And 3. The exploitation of that IP mutually developed. Here’s the takeaways of modern cost sharing: When we look at it intra from the transfer pricing aspects, we still have the inputs. We still have the outputs. How are we going to divide up the outputs?

But with the middle one, it’s really about who’s going to bear the risk. If we look from the 1966 original proposal on cost sharing regulations to the 2011 final to the OECD’s BEPS project on transfer pricing in general: who bears the risk, I think, is the most often forgotten aspect of the functional analysis.

On paper we may write that down, but – and this is where the rubber meets the road right now – we’re all suffering through the coronavirus. Many of you have supply chains in China. I remember about 10 years ago, a great case study in Europe now that was not great for Coca Cola.

The case study was, the French government decided there was a flu epidemic. I forget the exact illness, but it was a stomach virus and French government had decided it was Coke’s fault. It must be something to do with Coke. One could wonder if this was one of these French-U.S., you know, ‘tiffs.’ They said it was Coke’s fault and they shut down the Coke facilities and the Benelux. There was no Coke in France for a couple of weeks, no production from a transfer pricing perspective.

What became very interesting rubber-meets-the-road from a transfer pricing perspective: Coke had put on paper, at least all of its risk in Belgium, Belgium at that time had a tax incentive through a headquarters operation. And now one could look in that stress of that year of shutdown at who actually bore the risk.

Did that headquarters company report less earnings? Was it unable to make the distributions expected? Did it’s treasury management center soak up those losses or did the United States actually take deductions pass through the money to cover the losses? And the answer was Bennett Lux. The Belgium [entity] soaked up those losses, not because of transfer pricing, because that was the business operations, the functional analysis that Coke actually had.

I bring that up because now we have the coronavirus. Many of you all know that transfer pricing is an economic phenomenon. Forget the tax aspects. If we all had zero tax, if we all had zero tax regimes right now in the world, every country does away with tax somehow magically, we can fund government, we would still have transfer pricing!

Ronald Coase in 1930, you know, his seminal work about the firm said there’s reasons for firms to trade internally and not on the open market because of transaction costs and so forth. That’s why we have horizontal and vertical today.

We all know that practically speaking, but I think we forget where it comes from. So we’re going to have transfer pricing regardless in terms of intra-firm transactions. And regardless of if there’s tax, we’re going to have disputes along the lines of “Are we going to have more jobs in China or more jobs in the United States?”, [or] “Is the manager in China going to have more of the earnings and get a better bonus than the manager in the United States?” These are political realities, nothing to do with taxation. Now we have the coronavirus and it’s going to be interesting because rubber meets the road.

We’re going to be able to stress test: Where did companies say that their risk is and who actually bears the risk? Will the United States be making transfers to support, by example, supply chains in China? This is kind of a warning and a good thing!

Or will the Chinese operations of your companies be expected – like if they were in the United States to go out and find financing, maybe they’re going to have losses this year in China somehow – are they going to bear the risk that on paper they’re supposed to bear? If that’s where you have placed to risk on paper for transfer pricing purposes.

So the coronavirus is obviously not a good thing. It’s very bad for the world economy. However from a transfer pricing academic point of view, this is going to be a great year to stress test companies’ functional analysis of where they’ve placed that risk in their value chain.

Mimi Song: And then when you think about the idea of cost sharing, in some ways, we’re basically saying that if a company bear certain costs or would they ultimately also bearing the risk associated with [those costs].

William Byrnes: Absolutely.

Mimi Song: What I actually found interesting is you said you had talked about the history of cost sharing and going back to the initial cost sharing regulations, you had given me a little perspective here, and you said their original cost sharing regulations actually goes back to, say, that the risk to be born with respect to development activity is the possibility that such activity will not result in the production of intangible property – or  that the intangible property produced will not be of sufficient value to allow for the recovery of the costs of developing it. And I didn’t know that perspective!

And I thought that was fascinating because that is almost contrary to why people enter into cost sharing arrangements today? Like the idea is, “Hey, we’re going to share costs between related parties, because we’re thinking this is not really going to result in that much and we’re going to incur losses and we want to share those losses amongst our participants. Because in some way there’s perhaps the value to the organization, but not in monetary value. But yes, but yet cost sharing has evolved beyond that. It’s, it’s actually different from that initial perspective.

William Byrnes: So let me give you four reasons to enter into a cost sharing arrangement. One is Ronald Coase, his original view, right? First: If it doesn’t reduce transaction costs, then why are you doing it? But number two: how do we share the risk? Because the risk, as you know, I know somebody in the room is from pharma, but generally when we start setting transfer pricing, we start with pharmaceutical industry and it costs about $1.6 billion to bring a drug to market.

Not only is there risk, but then a third point of cost sharing: There’s the intangibles, but there’s the innovation that biotechs — by example, small biotechs that certainly don’t have 1.6 billion, they have the innovation, but they don’t have the capital or the experience to bring it to market.

We’ve all heard “Wow, that guy’s really smart, but he’s not a business person.” Or, “She’s a great professor, unfortunately she can’t monetize what she knows.” Therefore she’s an academic, but we’ve all heard this terminology.

So it’s risk bringing together partnerships and innovation, reducing or sharing costs, but also reducing transaction costs. So we have these different principles of why we come into cost sharing. But again, I think the one that’s most forgotten is about bearing risk.

And what does risk mean? It means the failure of production of the intangible property – and that was recognized in 1966. I don’t mean to harp on my friends in Treasury, but let me harp on Treasury: Why is Treasury forgotten that risk going into cost sharing means the sharing of the failure to develop intangibles or that the intangibles don’t turn out as expected?

And why do I bring that up? Let’s take the Amazon case. Who here knew back when Amazon started to transfer its intangibles to Luxembourg – for legitimate business reasons to develop out that world market, to do the language and so forth – that Amazon would be the biggest company in the world? That Jeff Bezos would be the richest man in the world? Who made that bet back then?

I remember at that time, Amazon was in operating losses. When at any definitive point did it look like Amazon would start making money? In fact the press reports were “What kind of company is this that has such a valuation at that time (not yet on the market though,) that people were looking at is the ‘Next Big Thing’ that couldn’t make money and had moved foreseeable opportunity to make money?” “Who would make that bet?” That from a cost sharing point of view is the perspective that Amazon entered into a third-party arrangement with Amazon Luxembourg.

And we have to stress that third party arm’s length arrangement. That’s our law — for the lawyers in the room: We live by a book. And the book was written by U.S. Treasury in 1935 for good geopolitical reasons because of tax treaties. Because of the United States being a capital exporting nation for good geopolitical reason said, “We want to go by the arm’s length theory as if third parties were transacting. When we divvy up the earnings pot between a foreign country and the United States.”

This was really the U.S. position, not all countries in the world. I promise you in 1935 when that regulation was written – but even before when it was being discussed at what is now the United Nations — when it’s being discussed today in the context of tax treaties, a lot of countries didn’t do it.

Internally, we don’t do that. Right? Internally, within the United States for state and local tax, we use formulary apportionment – which is not easier, which is not simpler, and will not produce a one-page tax form-system like with supposedly coming from the Tax Cuts and Jobs Act.

Let’s just bring it forward from 1966 to where we’re at today, U.S. Treasury seems to be bringing cases that are post facto, even when its own regulations and its own statement said that anticipated benefits from commensurate with income 1986 and the Section 482. And so we reasonably anticipated benefits, but those would be forward looking from the point of time of the agreement, but yet every case seems to be a post facto case.

It seems to be, we don’t like the result of [the Medtronics case]. We don’t like the result of the agreement that we did with Medtronics. The closing agreement produced better than expected results for the Puerto Rican subsidiary. So let’s challenge Medtronics.

There was the APA case with Eaton:We have an advanced pricing agreement. You know what? We don’t like the results. So let’s bring a case!”

And then the excuse? Well, Eaton said they got bad data, got bad results. They still came to treasury and said, “We’ve discovered we have some bad data. We had some bad results. We’re amending our return.” And Treasury was like, “This is our opportunity to tell you that we don’t like post facto the results. We have to redo the agreement.”

And Eaton said, “But we live by a book in that case, an APA or closing agreements – we live by these words in this piece of paper, this is how we made business decisions. Whether we would do a deal with you because it would be cheaper for litigation, whether we would do a deal with you because it would be better press. We can’t do this deal because it’s against the rule of law.”

There’s something fundamental as Americans that we want to be able to rely on what our government tells us.

Mimi Song: Well, it’s interesting going back to your point about Amazon, that there was actually business rationale for them being able to enter into cost sharing arrangement with Luxembourg. Why is it that multinationals tend to enter into cost sharing arrangements with low tax jurisdictions?

William Byrnes: That really is the elephant, isn’t it?

Mimi Song: It really is! [Laughs.] It just happens to be advantageous from a tax perspective. So let’s be honest. Right?

William Byrnes: Alright. Where did cost sharing and the transfer pricing disputes really start? It was really started with Puerto Rico, right? There was our Cayman’s case.

There was an oil case from Venezuela. It really started with Puerto Rico. The first pharmaceutical operation in Puerto Rico is recorded by department of commerce. It was 1960, the Puerto Rican tax credits and all that came about from the United States federal point of view much later in time, the late seventies.

By the time those came about, there were over 50 operating pharmaceutical manufacturers in Puerto Rico already. They went to Puerto Rico – and I actually know this because again, my mentor was Walter Diamond and Walter Diamond informed me of what actually happened. They were under the red scare in the 1950s, the missiles were going to come raining down at any time. This is what people believed and we needed medical pharmaceuticals and equipment to be able to revive, literally repair the body of the country.

And so we had to diversify the risk of our operations. And where was the American colonies? Well, we didn’t have any, America didn’t have colonies. So where was America outside of America? That the missiles wouldn’t rain down on? Puerto Rico is literally the only place. Guam was too far away and they didn’t have any infrastructure. Puerto Rico didn’t have any infrastructure, but that was the actual reason that pharmaceutical started to look at Puerto Rico, not because Puerto Rico was outside the U.S. tax system. And back in the late fifties and sixties, you’ll remember, there was no controlled foreign corporation legislation. But it wasn’t for low tax, it was for a diversification.

So let’s fast forward to today. It’s a Cayman Islands? Is it a U.K.-through-Bermuda?

Mimi Song: Ireland..?

William Byrnes: Ireland-to-Bermuda? Is it, is it just the low tax? Well, when I speak with my friends at Treasury, and again — I’m not always contrarian to them, but you know, if they’re supporting treasury then as an academic I feel I have to be on the other side. And as the contrarian, my question to them is “So tax shouldn’t be considered in the business decision making process. Do you consider tax not a cost?” and of course [they say,] “No, tax is an allowable cost, but this is unreasonable…”

And I said, “But when you say it’s unreasonable, do you mean by your personal perspective or do you mean I’m going to find that in the law of the United States that Congress, or even the United States Treasury, has written?” Because when I read what Congress has written that has been promulgated through a democratic system with lots of lobbyists, lots of dollars. But those dollars in lobbyists support our — I’m going to say this and I won’t explain it right now – they support our retirement because all the companies in the room are public. It’s my retirement counts in my state pension, or my 401k four, 3B that invests your companies.

But those laws were promulgated democratically. And those laws allowed those cost sharing agreements. And the IRS wrote that 1935 arm’s length [law] and the IRS in every iteration, including the 2009 proposed and 2000 final [regulations] said, “These regulations comply with the arm’s length standard.”

Had the IRS come out straight forward and it just said, “Arm’s length standard isn’t going to be applicable anymore for transfers of intangibles. It doesn’t say the word arm’s length in [Section] 482. It doesn’t say the word arm’s length in 367-D, Congress didn’t say arm’s length. Treasury told you arm’s length because of geopolitical reasons.

If Treasury had come and said, “You know what, we’re going to formulary apportionment going forward. Let’s make regulations. We’ll take the input, but that’s why it’s going to work…” But that’s not what Treasury told you.

Now these aren’t exactly third-party transactions. We know that from our Altera’s case where everybody in industry, every single group in industry came with documentation and all the other technology cases saying that arm’s length parties do not share in those cases. The big issue was the stock-based compensation: they don’t share it.

They said, “U.S. Treasury show us one agreement [where companies share stock-based compensation at arm’s length]” and U.S. treasury would say, “Well, the regulations comply with the arm’s length standard. The arm’s length standard is actually… I know you didn’t understand the arm’s length standard until now, but it’s actually an income-based method. It’s an asset purchase agreement-based method. It’s a…”

And then you have all the people who believed in the arm’s length standard saying, “But that’s not what the arm’s length standard has been for the last 80 years as you’ve told other geopolitically.”

I think the reason we don’t have that written in the regulation because Treasury could write it. And if they wrote it, I’d be like, “Yeah, that’s the rule.” Y’all would be “That’s the rule,” but you keep reading the arm’s like standard.

The reason they don’t do it because we want that to be the rule for relative to U.S. companies. But we don’t want it to be the rule for India to U.S. companies. We don’t want it to be the rule for China or Indonesia or Kenya for U.S. companies. And therefore we say to the world, “It all complies with the arm’s length standard,” but internally *nudge, nudge, wink, wink.*  We know it does it. And post facto [Treasury is telling companies,] “You’re getting a little too much. Let’s bring some of it back up.”

Congress can sort it out. Congress did! We have the Tax Cuts and Jobs Act, GILTI, FDII, BEAT… these new terminologies, these new acronyms … [some laughter in the room…]

Mimi Song: Very, very suitable by the way…

William Byrnes: These acronyms are very suitable, but Congress can redefine the word in tangibles as they did in 367 and 482. This is, for me, the most interesting aspect of Altera[‘s case]: It’s not the constitutional challenges of the Administrative Procedures Act. And it’s interesting, academically speaking, but here’s really the fundamental crux: What does the word intangible mean?

And it comes down to the word “any”. When you read 482, if you’re looking at it right now on the computer screen, “any transfer…” yada-yada-yada intangibles, “any transfer … intangibles…” the Treasury argued any does not modify the word that it immediately proceeds and tangibles any intangibles. [Grammarians would argue,] “Actually it works somewhat like a preposition and it modifies…” [while Treasury is arguing,] “I’m sorry, ‘any’ transfer it, doesn’t modify the word ‘transfer.’ It modifies the word intangibles…”

Matthew DeMello: Just to interrupt quickly. Fiona, in 2014 legislation under the Obama administration expanded the scope of intangibles. Can you tell us how the definition changed?

Fiona: They’ve changed quite a bit. Actually the IRS wants to limit the reduction in the U S tax base from cost sharing arrangements, and they’ve changed regulations several times throughout the years to do so. First, there was the qualified cost sharing regulations issued in 1995. Those regulations tightened on stock-based compensation in 2003 mandating that they are included in cost pools. Temporary regulations were issued in 2009, then final regulations in 2011. You can see how cost sharing arrangements often end up in dispute.

William Byrnes: So intangibles can have it expansive meaning because Congress put ‘any’ here and put in tangibles at the end of the sentence, but meant for ‘any’ to be in front of the word intangibles. That’s Treasury’s argument. And the company said, “But the way I read English and the way I read statutes, not just the statutory construction but just the simple English, iff a word proceeds another word, it modifies that word.

It proceeds that’s Congress could have written it differently. The people who write our tax laws are you, when you’re younger, working as a staff member on the Hill. It’s you as a lobbyist who are seeking to influence for the benefit of your company, which benefits America, which benefits my retirement account. So I trust your lobby dollars.

I do. I’m not one of those people says we shouldn’t have lobby dollars. I want you to representing my retirement account by representing your company because your earnings go to my 401k or my pension plan at Texas A&M.

I’ll stop with that. But I think it just opens up a different perspective for you to think about when we talk about cost sharing. Because what I read in the newspaper is always low tax countries, and it’s the way I interpret it: This is what Congress wants you to do. And if Congress doesn’t want you to do it, it makes GILTI, BEAT, FDII… Transparency on taxation. Congress can do a lot of things and it’s chosen not to.