A Battle Against Base Erosion: D.C.'s ‘Chainbridge' Saga

For over 50 years, Bloomberg BNA’s renowned flagship daily news service, Daily Tax Report® has helped leading practitioners and policymakers stay on the cutting edge of taxation and...

Transfer Pricing

In the second article of a two-part series, Bloomberg BNA looks at the obstacles confronting states as they combat abusive transfer pricing. This article focuses on the efforts undertaken by the District of Columbia to combat revenue losses through its contract with economic consulting firm Chainbridge Software LLC. Taxpayers fought back vigorously against the District's transfer pricing program and in 2012 scored a major victory when Microsoft Corp. successfully challenged a $2.75 million assessment based on a Chainbridge study. There an administrative law judge found that the “Chainbridge method” didn't comport with federal regulations. But the story is far from over. At least one former state official says it isn't clear that the judge got it right. Meanwhile, an appeals court has remanded three other cases to the Office of Administrative Hearings, cranking up pressure on taxpayers to settle rather than fight.

By Dolores W. Gregory

Aug. 30 — Economist Eric Cook recalls the turning point in his thinking about transfer pricing. Something he had thought of as a mere accounting mechanism for related-party transactions had been co-opted by multinationals as a tool for tax avoidance.

His epiphany came in the middle of a meeting with senior tax officials in Alabama, late in September 2001, as he and his wife and business partner, Nancy, pored over revenue figures that had them scratching their heads.

The Cooks' consulting firm, Chainbridge Software LLC, offered revenue modeling services to states, and officials at the Alabama Department of Revenue wanted their opinion about the numbers.

“They had passed tax legislation that unequivocally should have been an increase. Their economy was doing very well at the time. But they had a decline in revenue,” says Cook. “And that's when it occurred to me: This was transfer pricing.”

Since that day, Cook's epiphany has hardened into an almost evangelical zeal to level the playing field for small states, whose revenue departments are often too overwhelmed to handle such complex audits on their own.

This work has landed the Cooks in the thick of a legal controversy in the District of Columbia, where their transfer pricing analyses of multinationals such as Microsoft Corp. and BP Products North America have drawn fire. So far, the results have been mixed—a victory for the District in the case of BP Products, and a stinging defeat in the Microsoft case.

For now, the D.C. transfer pricing program is on hold, pending the outcome of three other cases. ExxonMobil and two other oil companies are challenging $3.8 million in assessments based on transfer pricing studies by Chainbridge, making the same arguments as Microsoft.

Depending on the outcome, those cases ultimately could vindicate the Cooks. Or they could convince other states that it is simply too risky to engage them.

‘Bringing a Knife to a Gunfight'

In many ways, the Chainbridge story illustrates the overwhelming challenges facing states that take on aggressive tax planning by multinational corporations—particularly when it involves a mechanism as complex and imprecise as transfer pricing (168 DTR J-1, 8/30/16).

“It's a classic case of bringing a knife to a gunfight,” says Billy Hamilton, executive vice chancellor and chief financial officer with Texas A&M University, who is a former Texas deputy comptroller of public accounts.

State revenue departments have nothing close to the kinds of resources of Big Four accounting firms and the major international law firms that represent corporate taxpayers, he says.

“The states are understaffed, and these auditors—and some of them are incredibly good—but a lot of times, if they are any good, they are hired away almost immediately,” Hamilton says. “So states are undermanned to do what they are called on to do.”

Chainbridge attempts to offer an affordable solution. It analyzes corporate profit margins (CPM) using a proprietary software that crunches tax return data supplied by the states and develops a list of companies whose profits appear to be suspiciously low for their industries. Then the company performs more detailed studies of the taxpayers that the client—the state—wants to audit, using another software program that automates some of the economic research.

“That's really the bottom line here,” says Nancy Cook, Chainbridge's president and primary transfer pricing specialist. “We're actually just trying to help these guys get money for their states. And there's an army of transfer pricing people on the corporate side. We estimate something like 13,000 professionals worldwide. And so the states are outgunned.”

‘Magic Eight-Ball' Method

In court filings and published articles, attorneys for taxpayers in the District of Columbia have derided the “Chainbridge method” of transfer pricing analysis as a “Magic Eight-ball” approach that relies on numbers pulled from the ether to justify outrageous income adjustments. They have argued that the method is arbitrary and capricious because it fails to segregate related-party from unrelated-party transactions.

The taxpayer's argument persuaded Administrative Law Judge Paul Handy to strike down a $2.75 million franchise tax assessment against Microsoft in 2012. But it didn't convince D.C. Superior Court Judge John Campbell to do the same in a case involving BP Products North America.

Finding too many facts in dispute, Campbell scheduled the BP Products case for trial in early 2014. BP settled the case in January that year, agreeing to a reduced assessment of $580,000 ( BP Prods. N. Am. Inc. v. District of Columbia, D.C. Super. Ct., No. 2011cvt10619, settlement 1/20/14 )  (19 DTR K-5, 1/29/14).

For the District of Columbia Office of Tax and Revenue (OTR), BP Products was a clear victory. For the Cooks, it wasn't an ideal outcome. Though Campbell seemed to appreciate the District's position, he never issued a written opinion that could have countered some of the damaging conclusions Handy drew in the Microsoft ruling.

Mystifying Decision

The Cooks say they are mystified by the taxpayers' arguments and astonished by Handy's ruling that the “Chainbridge method” of analysis “failed to comport” with federal transfer pricing regulations under tax code Section 482.

“There is no ‘Chainbridge method,' ” says Nancy Cook. “We follow the regulations guiding the application of Section 482.”

One former state official agrees that the Microsoft ruling is a head-scratcher.

“It was clear the administrative law judge had no understanding whatsoever of how to do transfer pricing,” says James Wetzler, former commissioner of taxation and finance with New York state. Wetzler also was a director in Deloitte Tax LLP's New York City office, retiring in 2013 after nearly 20 years with the firm.

One can question whether Chainbridge used the right method to analyze the company's transfer pricing, Wetzler says, and one can argue that Chainbridge took shortcuts in getting there. But Handy didn't really address those questions.

Instead, he conflated the requirements of a transactions-based method of analysis with the requirements of the comparable profits method, and effectively held that the CPM itself was wrong, Wetzler says.

“But the CPM is a valid method, and in most cases, ends up being the right method,” Wetzler says. “He didn't know what he was talking about. That happens when you litigate things. The judges can be clueless.”

Origin Story

If you ask the Cooks to describe Chainbridge, they will say it is a software company specializing in support services to state revenue departments. Transfer pricing is only one line of business. The company also provides revenue modeling services to states, to help administrators forecast the impact of changes to tax policies or tax rates.

Revenue modeling is Eric Cook's speciality. It was the focus of his work at the Joint Committee on Taxation, where he went to work after completing his doctorate in economics at Florida State University in 1983.

That same year, Nancy Cook graduated from Pennsylvania State University with a bachelor's degree in British literature—“Shakespeare, Milton, Chaucer”—and an ambition to move to Washington. Their paths crossed a few years later at the former Price Waterhouse, where Nancy had worked her way from a secretarial to a professional track, acquiring a Master of Business Administration from George Washington University along the way.

She found she had a head for numbers and a growing fascination with computers.

“I would come in on the weekends to do spreadsheets, just for fun,” she says. “I loved it.”

Her former boss, economist John Simpson, recalls her as a dogged and talented researcher with a flair for writing clear analysis.

“It was a good schoolyard for her, a good training ground,” Simpson says.

He recalls that she was especially adept at running down data for comparables. In those early days, before the existence of the databases that are in wide use now, running down comparables meant a trip to Securities and Exchange Commission offices to page through paper copies of company Form 10-K annual reports.

Inspiration for Software

It was tedious work that Nancy remembers as “almost painful.” But that labor became the impetus to develop ways to automate portions of the research. One of the first programs she developed at Price Waterhouse was something she called “TP Test.”

“The idea behind it was to generate consulting revenue in our D.C. transfer pricing practice,” she recalls. “It was a little floppy disk. The idea was we would sent it to the practice offices, and they could enter data for their clients and then run a test—and it would go against the Compustat data and you could see if they might have a transfer pricing issue and then the idea was they would call us in D.C. and we would get transfer pricing consulting work.”

That never quite got off the ground, Eric says, because the transfer pricing specialists were worried that accountants in the field would try to do the transfer pricing on their own. But it was the genesis of the preliminary analysis software that Chainbridge eventually developed and uses now.

In 1992 they moved on to the Barents Group, then part of KPMG, where they met software designers Kirk Copen and Jamie Copenhaver. Co-founders of Chainbridge, Copen and Copenhaver developed the software the company uses to conduct its transfer pricing studies.

Crunching the Numbers

In Chainbridge's modest offices in Fairfax, Va., Eric Cook fires up his laptop computer to demonstrate the software program that the company patented in 2010.

Using tax returns provided by the state, Chainbridge compares taxpayers' profit margins with those of comparable companies within the same industries. To run the comparisons, they pull in comparables from Compustat, a database of financial, statistical and market information for global companies developed by Standard & Poor's Financial Services LLC.

The analysis looks at the ratio of operating profits—net profits divided by sales—and flags those taxpayers whose margins seem unusually low in comparison to the comparables, Copen explains. “If you have a state with 40,000 corporate tax returns, there are probably three or four hundred that get adjustments—off the preliminary adjustments—and then one or two hundred that have adjustments worth looking at.”

Chainbridge then narrows the list by looking at the companies' public financial data—such as annual reports filed with the SEC—to see if the taxpayers reporting losses for tax purposes are also reporting losses to their shareholders. Those that are will be dropped, Copen says. Those that aren't will be flagged.

“It is a small fraction of taxpayers, but a lot of times the ones that are in violation can have significant amounts. You might have a state where you've got two to three hundred million in tax adjustments and the bulk of that is maybe spread across 75 to 100 companies,” Copen says. “That, I think, is not unusual.”

Says Nancy Cook: “This is the thing we have the patent on. This software, this process. This is not the final transfer pricing analysis. This is just to find possible candidates, but I think people somehow think that we just spit numbers out and don't think about them.”

Good Analytics

States use a similar kind of analysis, Hamilton says, to root out Medicaid fraud, by identifying patients who ping-pong from one doctor to another.

“It's the same technology,” Hamilton says. “It's good analytics. It points you in a direction more than gives you an answer.”

When Hamilton was working as Texas deputy comptroller, he says, he met with the Cooks on several occasions to discuss using their services. Nothing came of it, he recalls, partly for budgetary reasons and partly because Texas at the time imposed a single entity franchise tax that was unlike most other corporate income taxes.

But he says he isn't surprised that states have gotten pushback from taxpayers over the use of Chainbridge's services. “They are basically doing an analysis that sets people up for a tax that their accountants say they don't owe. That's always going to be a fight no matter how valid it is.”

Program in D.C.

In the District of Columbia, revenue officials have long been aware that the District's corporate income tax base was eroding.

“It is pretty well documented that there has been transfer pricing going on for years,” says Glen Groff, director of operations with the OTR.

Many of the transfer pricing issues were related to Delaware holding companies, he says—structures organized solely to hold intangibles, such as trademarks, patents and copyrights, and lease them to operating companies in higher-tax jurisdictions. For the holding company, the royalty payments are untaxed income. For the operating company, they're a deductible expense.

For the states where those operating companies are housed, those royalties mean lost revenue.

To stem the tide, the District adopted an “add-back” statute in 2004 to disallow certain deductions for intercompany payments. But Groff says the statute wasn't very effective, at least initially.

Chainbridge offered a more promising solution. In 2008, the District launched the program, contracting with the Cooks to review 24 tax periods each year. That translates to audits of eight companies—out of some 30,000 corporate taxpayers—because each audit involves a three-year rolling average, Groff explains.

The District has been pleased with the results.

“It was actually very successful with a number of companies,” Groff says. “Obviously the initial assessments we made were eventually adjusted, but most of them identified a shifting of money within the company that resulted in a shortfall for the District.”

Functional Analysis

To prepare the transfer pricing studies, Chainbridge draws information from the company's tax returns—provided by the state—as well as publicly available data, and to the extent the Cooks can obtain it, the taxpayer's response to their functional analysis questionnaire.

In any transfer pricing report, the functional analysis is the heart of the study. It is where the significant activities undertaken, the risks assumed, and the assets and resources employed by the controlled entities that are the subject of the study are identified. If the company has sales to unrelated parties, those sales might serve as internal comparables for benchmarking intercompany prices. The absence of transactional data can point the analyst toward a profits-based approach.

Chainbridge has come under criticism from taxpayers, as in the D.C. litigation, because it defaults to the comparable profits method. It generally doesn't look at specific transactions. And there is a reason for that, the Cooks say. Taxpayers never turn over the data they need to conduct that kind of analysis.

“We'd be happy to perform the analysis using other methods, but they never provide us with the information,” Nancy Cook says. “We ask them to give us the general ledger information. And they won't do it.”

Only once, out of the nearly 90 audits Chainbridge has worked on over the years, has a taxpayer provided transactional data, she says. Often, however, taxpayers will produce more detailed information once a proposed assessment is issued. When they do, Chainbridge takes a second look.

That's what happened in the BP Products case.

“The taxpayer had an issue with the way that exploration and production companies do their accounting,” she says. “So we revised the adjustment and came down.”

BP Products originally paid the $720,000 assessment and sued for a refund in the D.C. Superior Court. On review, Chainbridge reduced the assessment to $580,000, which BP agreed to pay. The company walked away with a $140,000 refund.

“That's an example of where the taxpayer came back with information and an opinion and we think about it and we say we agree,” Nancy says.

Reliance on One Method

The fact that Chainbridge relies almost exclusively on the CPM is a problem in Wetzler's view, because it means the company starts out taking a shortcut.

“He doesn't do a best methods study,” Wetzler says. “He does the CPM for everybody. Most of the time the CPM is the best method, but not all the time.”

Wetzler has known Eric Cook since they worked together at the Joint Commission on Taxation in the 1980s, when Wetzler was chief economist and Cook a revenue estimator.

“He did a good job,” Wetzler says of Eric Cook. “He's a smart guy who gets stuff right more than he gets stuff wrong.” But Wetzler says he also believes the Cooks are constrained by what states are willing to pay—which isn't nearly enough, in his view.

“To do a really high quality transfer pricing study, you have to subpoena a whole bunch of information,” Wetzler says. “That means doing a laborious study.”

The Internal Revenue Service conducts transfer pricing audits routinely. In major cases, it will issue dozens of information document requests (IDRs), conduct multiple site visits and interview a number of company officials. In one particularly contentious audit—of Eaton Corp.—revenue agents spoke to more than 80 employees.

When taxpayers decline to turn over information, the IRS frequently issues summonses and goes to court to enforce them. The agency stirred up a storm recently when it contracted with an outside law firm to assist in an audit of Microsoft's transfer pricing. The bill for the first phase of the contract alone was $2 million.

Chainbridge, in contrast, sends out a single IDR to taxpayers. They don't go into the field, and the Cooks say no state ever issued a subpoena for information that taxpayers refused to provide. What Chainbridge conducts are essentially desk audits.

“States are not prepared to spend the money necessary to do a full-fledged best methods analysis,” Wetzler says. “Eric is doing the economy version of what you would do.”

But at the same time, Wetzler says, in many cases the Cooks are finding real trouble. There is abusive transfer pricing going on.

“If a company loses money for 15 years in a row, they are either running an extremely poor business or they are manipulating the transfer pricing,” Wetzler says.

Chainbridge Responds

The Cooks say they disagree with the characterization of their work as less than a “full-fledged analysis.”

“When we first started doing this work over a decade ago, we requested quite a bit of information from the taxpayers in the IDR, including copies of federal tax returns and the accompanying worksheets, transfer pricing studies, and APAs,” they say in an e-mail to Bloomberg BNA.

“We also asked that the taxpayers respond to a Functional Analysis Assumption Questionnaire (FAAQ) that covered fifty to seventy-five items, as we tried to not be too burdensome. Once we became involved in litigation in the District we began requesting more information (specifically, transaction-level data from the general ledger and a FAAQ that quite often is over one hundred pages long).”

And a big price tag isn't necessarily a guarantee of better quality, they say.

“We grew up in the accounting and consulting firms and we understand why their analyses are so expensive. We are not driven by the chargeable hour and are quite efficient in performing our work in that we are a software company and have been able to automate many of the processes that the accounting and consulting firms do ‘by hand.' ”

In the course of their work for several states, they say they have reviewed several hundred transfer pricing studies prepared by accounting and consulting firms. And they are struck by two observations: “Many of them look like our reports,” they say. “We also note that nearly all of these studies employed the CPM.”

Approaching Reasonableness

Groff says he isn't bothered by the criticism that the OTR's transfer pricing program has led to inexact adjustments.

“Audits are pretty dynamic animals,” Groff says. “They are not hard and fast. And the larger you get—with a large corporation—you're looking at samples of documents. It's not as simple as it would be for an individual where you look at every document.”

He adds: “A full-blown audit of a multinational would be almost impossible to do. It's not a perfect world, but it gets to the area of reasonableness.”

But Wetzler says he isn't so sure. Even if Chainbridge is identifying genuine problems in transfer pricing, he says, the big question is whether their studies will hold up in court—especially if a taxpayer with deep pockets is willing to pay a Big Four accounting firm the half million dollars or more required to conduct a more thorough transfer pricing study.

“Microsoft decided to challenge it,” Wetzler says. “And because of the crazy decision, all of the other companies got emboldened to fight.”

ExxonMobil Cases

Currently, more than half a dozen multinationals have appeals pending in the D.C. Office of Administrative Hearings (OAH) over transfer pricing assessments.

‘Chainbridge' Transfer Pricing Cases in D.C. Case Issue Status AT&T Services Inc. v. D.C. Office of Tax and Revenue, OAH Docket No. 2014-OTR-00026, filed 9/12/14. An operator of wireless services challenged a $4.3 million assessment for franchise taxes based on a transfer pricing study conducted by Chainbridge. Stayed pending the outcome of the district's appeal in three other cases involving ExxonMobil, Hess Corp. and Shell Oil Co. BP Products North America v. District of Columbia, D.C. Superior Court, Docket No. 2011cvt10619, settlement in principle, 1/28/14. The company sought a refund of $722,585, an assessment based on a transfer pricing study conducted for the district by Chainbridge. BP Products also sought 6 percent interest from the date the assessment was paid until the date of a refund by the D.C. Office of Tax and Revenue. D.C. Superior Court Judge John Campbell denied the taxpayer's motion for summary judgment, finding that there were factual issues that could only be decided at trial—such as whether the proper transfer pricing method had been applied. The parties then came to a settlement agreement, under which the district retained $581,600 as satisfaction of the assessment, and BP Products received a refund of $140,985 plus 6 percent interest from April 6, 2011, to the date of the refund payments. D.C. Office of Tax and Revenue v. ExxonMobil Oil Corp., D.C. Court of Appeals, Docket Nos. 14-AA-1401, filed 12/12/14. Exxon and two other oil companies challenged multi-million-dollar assessments for tax years 2007-09 based on transfer pricing studies conducted by Chainbridge. A previous administrative case involving Microsoft Corp. was decided in the taxpayer's favor, but OTR maintains the decision has no precedential value for similar cases. The case is combined with two others involving taxpayers Hess Corp. and Shell Oil Co. Administrative Law Judge Beverly Sherman Nash granted the taxpayers' motions for summary judgment 11/14/14, finding that the district is bound by the Microsoft ALJ ruling. The D.C. Court of Appeals overturned Nash's ruling June 30 and remanded the case to the Office of Administrative Hearings for further consideration. Eli Lilly and Co. v. D.C. Office of Tax and Revenue, OAH Docket No. 2014-OTR-00024, filed 9/5/14. An international pharmaceutical company challenged a $506,000 assessment for franchise taxes based on a transfer pricing study conducted by contractor Chainbridge. Stayed pending the outcome of the three other cases involving ExxonMobil, Hess and Shell Oil. Exxon Mobil Corp. and ExxonMobil Oil Corp. v. D.C. Office of Tax and Revenue, OAH Docket No. 2014-OTR-00025, filed 9/12/14. Two separately incorporated entities of the oil conglomerate challenged an assessment for $876,000 in D.C. franchise taxes based on a transfer pricing study conducted by Chainbridge. The taxpayers note that the assessment was issued to Exxon Mobil Corp. using the federal employer identification number of ExxonMobil Oil Corp. Stayed pending the outcome of three other cases that involve the same issue. Microsoft Corp. v. Office of Tax and Revenue, OAH Docket No. OTR-2010-00012, decided 4/23/12. A District of Columbia administrative law judge struck down a $2.75 million assessment against Microsoft Corp., finding that the transfer pricing analysis on which it was based was useless in determining whether Microsoft's controlled transactions were conducted in accordance with the arm's-length standard. The District of Columbia dropped its appeal of the case. Pfizer Inc. v. D.C. Office of Tax and Revenue, Docket No. 2013-OTR-00012, filed 4/30/13. Pfizer is challenging a $1.7 million assessment for tax years 2007-09 based on a transfer pricing study conducted by Chainbridge. Administrative Law Judge Mary Masulla granted OTR permission to conduct discovery. Siemens v. D.C. Office of Tax and Revenue, OAH Docket No. 2013-OTR-00010, settlement July 16, 2013. Siemens's transfer pricing dispute wasn't publicly disclosed until the settlement was announced after a hearing on the ExxonMobil cases in November 2014. The amount in dispute wasn't disclosed, nor was the amount of the settlement.

Most are stayed pending the outcome of the three consolidated cases involving ExxonMobil Oil Corp., Shell Oil Co. and Hess Corp. Those taxpayers argued—successfully—that the Microsoft ruling established that Chainbridge's transfer pricing studies are invalid as a matter of law.

Attorneys with McDermott Will & Emery, who also represented Microsoft, have said that their clients should be able to rely on Handy's ruling in that case. Citing the doctrine of non-mutual collateral estoppel, they persuaded ALJ Beverly Sherman Nash to strike down some $3.8 million in assessments against Exxon, Hess and Shell that were based on Chainbridge's work.

Collateral estoppel essentially restrains parties from relitigating an issue that has already been decided. But the doctrine is almost never applied in tax cases, where different facts and circumstances drive to different results. Alan Levine, chief counsel with OTR, told Bloomberg BNA that OAH decisions are never considered to have precedential value, except for the taxpayer directly affected by the ruling.

A ruling to the contrary would have profound implications for the District, says Stephen Cordi, deputy chief financial officer with OTR.

“As far as the District is concerned and our attorney general is concerned, this issue has a far wider application than just taxes,” Cordi says. “Essentially it says to the District that if you fail to appeal an adverse administrative decision, it's binding on you forever. So your attorney general can't say he'd rather litigate with better facts. That forces the attorney general to appeal almost any decision whether he likes it or not.”

In the District, a taxpayer can challenge an assessment in OAH or pay the tax and sue for a refund in superior court, as BP Products did. For both forums, the venue for appeal is the District of Columbia Court of Appeals.

Cordi says the District made a strategic choice not to appeal the Microsoft ruling because “it had some unusual facts.” District officials expected that another case would present a better opportunity to argue for the government's position. However, that strategic calculation opened the door for taxpayers to argue that the Microsoft decision had settled once and for all the question of whether the District could rely on Chainbridge's transfer pricing analyses.

Appeals Court Reverses

But it appears that the question isn't yet settled.

The District of Columbia Court of Appeals June 30 vacated Nash's order and remanded the oil company cases for further consideration. The appeals court found that Nash had failed to explain the extraordinary circumstances that would justify the application of non-mutual collateral estoppel against the District.

Though the reversal was made on technical grounds, it raises the possibility that those cases eventually will be heard on the merits, one by one. If that happens, Wetzler says, then there is a good chance that the companies, like BP, will fall in line to settle with OTR.

Regardless of the outcome, however, it isn't clear whether the District will resume its transfer pricing program, Cordi says.

“We would have to see what the situation is. A lot of water's gone under the dam since these cases first arose,” he says. There have been some significant changes in District law, including a stronger add-back statute and the adoption of combined reporting.

In the meantime, the Cooks say they are hopeful that they will eventually be given a chance to speak in their own defense.

Because the Microsoft case was decided on a motion for summary judgment, and BP settled before it went to trial, they never had the opportunity to testify.

“It's definitely frustrating to have the thing decided by someone who probably doesn't have any transfer pricing experience, and you aren't allowed to say anything,” Nancy Cook says.

If the cases are heard on the merits, that means the Cooks may finally take the stand. And when that happens, Nancy Cook says she knows exactly what she would like to say.

“My background, my training—I do know what I'm doing,” she says.

No Priority of Methods

And she says she would tell the ALJ that there is no priority of transfer pricing methods under the federal regulations. She would explain that the CPM is set out in the federal rules, and that rules for the CPM don't require related-party transactions to be separated from other transactions.

She would explain that doing so makes no sense, because it would distort the results.

And Eric Cook says he would explain that Chainbridge isn't wrong in its choice of the tested party. Despite what the taxpayers argued in the Microsoft case and BP Products, it was entirely appropriate to focus on the state taxpayer as a whole and not on any related party that might be on the other side of a set of sales or purchases.

The whole purpose of the exercise, after all, is to determine whether the state has apportioned to it the appropriate share of a company's U.S. earnings.

It is true, he says, that under the CPM, the objective is to define the business activity as narrowly as possible, so that the activity under study encompasses the controlled transactions to the exclusion of most uncontrolled transactions.

But the regulations also make clear that “sometimes you just can't do it,” Eric says. “Because you might have a manufacturer pulling in components from related parties, pulling in components from unrelated parties, putting those components together to manufacture a product—and the final product will have both related party components and unrelated party components and there is no way to segment that. And what we find is this is how it works in Microsoft, this is how it works in BP, this is how it works in Exxon, Shell, Pfizer—all these companies, it's the same thing.”

Years ago, it might have been different. Years ago it was easier to isolate just the controlled activity—because it was often housed in a dedicated subsidiary such as a Delaware holding company. But the transfer pricing game has changed as multinational corporations have become more and more complex.

“They are global companies, with subsidiaries, a hundred, two hundred subsidiaries all over the world,” Eric says. “So they don't really have to get to a structure where it's based on tax planning. They can tweak their transactions so it moves money to where ever they prefer.”

Wetzler says he wonders whether states can ever really get to a winning position when it comes to international transfer pricing.

The IRS has thrown a lot of money at big transfer pricing cases, he says, and it keeps losing them in the U.S. Tax Court.

“Everybody thinks they're being taken to the cleaners by multinational corporations,” Wetzler says. “But if the IRS with all its resources can't do it. I'm skeptical that the states can.”

To contact the reporter on this story: Dolores W. Gregory in Washington at dgregory@bna.com

To contact the editor responsible for this story: Ryan Tuck at rtuck@bna.com

Copyright © 2016 The Bureau of National Affairs, Inc. All Rights Reserved.