GlaxoSmithKline and the IRS Finally Find Relief with Zantac
Longstanding Dispute over Transfer Pricing Settled

By Sharon Burnett and Darlene Pulliam

E-mail Story
Print Story
JUNE 2008 - When it comes to transfer pricing, even the most seasoned tax professionals can find themselves with a stomachache. Transfer pricing is at the heart of a long-running dispute between the IRS and a corporate taxpayer, which resulted in the largest settlement in IRS history. For 14 years, GlaxoSmithKline (GSK) and its predecessor companies had disagreed with the IRS about the transfer prices the U.S. subsidiary of GSK paid its U.K. parent for several different drugs. The dispute primarily involved the appropriate pricing method for GSK’s Zantac, a product designed to treat stomach ailments and ulcers. On September 11, 2006, GSK and the IRS agreed to a $3.4 billion settlement, the largest in IRS history.

Despite its magnitude, the settlement, did little to help multinational companies with setting inbound transfer prices when intangible assets are involved. The settlement does send a strong message that the IRS is very serious about taxpayers complying with transfer pricing rules. A closer look at the 14-year disagreement between GSK and the IRS may be instructive for taxpayers in similar situations.

Transfer Pricing

The primary issue in the GSK case was the transfer price at which the U.K. parent, GlaxoSmithKline PLC, sold drugs (primarily Zantac) to its U.S. subsidiary. In a parent-subsidiary transaction such as this, the transfer price does not affect the overall profit of the group, but it does affect the overall taxes paid by the group.

A transfer price should be equal to the price that would be charged in an arm’s- length transaction. This arm’s-length amount must be determined using one of the following methods:

  • The comparable uncontrolled price method (Treasury Regulations section 1.482-3);
  • The resale price method (section 1.482-3);
  • The cost plus method (section 1.482-3);
  • The comparable profits method (section 1.482-5); or
  • The profit split method (section 1.482-6).

The resale price method is approved by the IRS when a tangible product is involved. The resale price method is ordinarily used in cases involving the purchase and resale of tangible property in which the reseller does not add substantial value to goods that are distributed by physically altering them before resale. Packaging, repackaging, labeling, or minor assembly are ordinarily not considered physical alteration. The arm’s- length character of a controlled transaction is tested against the gross profit margin realized in comparable uncontrolled transactions. The resale price method measures the value of functions performed. Ordinarily, it should not be used if the controlled taxpayer uses its intangible property to add substantial value to the tangible goods [Treasury Regulations section 1.482-3(c)(1)].

An arm’s-length price is determined by subtracting the appropriate gross profit from the applicable resale price for the property involved in the controlled transaction. The applicable resale price is equal to either the resale price of the property involved or the price at which contemporaneous resales of the same property are made. The appropriate gross profit is computed by multiplying the applicable resale price by the gross profit margin earned in comparable uncontrolled transactions [Treasury Regulations section 1.482-3(c)(2)].

Advanced Pricing Agreement Between IRS and SmithKline Beecham

Very little information is available from the various advance pricing agreements executed by the IRS. The SmithKline Beecham advanced pricing agreement, which was obtained by Tax Analysts (Tax Notes Today, Sept. 18, 2006), was a major factor in the recently settled litigation between GSK (the successor entity) and the IRS.

The 1993 advance pricing agreement (APA) agreed to by the IRS and SmithKline Beecham governed intragroup sales of the drugs Tagamet and Dyazide. The document established a pricing method that the parties agreed was “consistent with the arm’s length standard” and achieved a result that “clearly reflects income.” In the APA, the company agreed to price Tagamet using the resale price method and the IRC section 936 cost-sharing payment. The resale price was to be calculated by reducing the company’s net trade sales for a 28% marketing commission and another 8% for the Tagamet trademark and the company’s trade name. As a result, the transfer price for Tagamet sold to U.S. parent SmithKline Beecham by SmithKline Beecham Pharmaceuticals Co. (LabCo.), a manufacturing concern in Puerto Rico owned by an SmithKline Beecham Corp. subsidiary, would be 64% of net trade sales. The agreement defined net trade sales as sales minus returned items, allowances, and any cash discounts that do not exceed 2% of sales (Tax Notes Today, Sept. 18, 2006).

General Transfer Pricing Example

Developing a drug and getting it to market involves numerous different costs, such as research and development (including patents), government approval, labor, raw materials, shipping, and marketing. The first two costs occur over numerous years and can be very difficult to track because companies are generally developing several drugs at a time. Exhibit 1 contains a possible breakdown of the costs (as a percentage of sales) of an inbound product with intangible components. The percentages in the example are loosely based on the APA between SmithKline Beecham and the IRS for Tagamet. The resale price method was used, and the APA was based on an outbound transfer price (see Exhibit 2 for a glossary of terms).

Case History

Allen & Hanbury Ltd., a U.K. research and development subsidiary of U.K. parent Glaxo, developed an ulcer drug during the late 1970s. Zantac was available for sale by 1981. A competitor, SmithKline, was able to get a similar ulcer drug, Tagamet, to market a couple of years earlier. In 1992, the IRS began an audit of Glaxo’s tax returns for 1989 and 1990. In 1994, Glaxo and the IRS attempted to resolve the dispute through the APA process (see the timeline in Exhibit 3).

At this time, the APA process was relatively new. It was difficult to imagine an agreement on a transfer price before a transaction actually occurred. Even though the purpose of an APA is to resolve transfer pricing issues before they arise during an audit, taxpayers and the IRS had already encountered numerous problems in agreeing on a transfer price after the transaction had occurred (e.g., Eli Lilly and Bausch & Lomb). The first APA was signed in 1991, and by 1994, a total of 45 had been completed. About half involved a foreign taxing authority. Taxpayers, most foreign taxing authorities, and the IRS were eager to agree on a solution to a transfer price. Between 1991 and 1994, about 49% of the received applications were executed. From 1995 to 1999, the rate increased to 61% (APA statistics from Announcement 2000-35, 2000-1 CB 922). The IRS has now reached the first APA with China (News Release 2007-09, 01/12/2007).

Unfortunately, Glaxo and the IRS were not able to come to an agreement in 1994. By 1999, GlaxoWellcome (GW) and the U.K. taxing authorities had reached an agreement on transfer pricing. In December 1999, GW requested relief from double taxation through the U.S.–U.K. Convention for the Avoidance of Double Taxation, commonly known as the “competent authority process.” (The transfer prices were important to determine taxable income in both countries.) The fundamental issue that remained after the competent authority process, the one needed to determine the inbound transfer price, was as follows: Which was more valuable, the research and development process completed in the United Kingdom or the advertising and marketing efforts in the United States?

The IRS asserted that the advertising and marketing were more valuable, because Zantac was the second ulcer drug to enter the market. GW and the U.K. taxing authorities believed a high transfer price was reasonable because the R&D was more valuable, and the U.S. GW subsidiary did not own any drugs. Ordinarily, taxing authorities, particularly the United States and the United Kingdom, are able to negotiate their differences. However, in this case a very large amount of disputed taxes was at stake—up to $5 billion.

In December 2000, GW merged with SmithKline Beecham Corp. (SKB) to become GlaxoSmithKline (GSK). This merger was important because SKB had been able to reach an agreement with the IRS over the outbound transfer price for Tagamet in 1993. As noted above, when the terms of the APA were made public, it was revealed that the outbound transfer price for Tagamet was 64% of the net trade sales (defined as sales less returned items, allowances, and any cash discounts up to 2% of sales). The 36% reduction from the net trade sales amount consists of 28% as a marketing commission, 5% for the Tagamet trademark, and 3% for the SKB trade name.

In May 2001, GSK and the IRS filed a joint motion to allow depositions—GlaxoSmithKline Holdings (Americas) Inc. v. Comm’r (117 TC 1). The IRS had not yet filed a notice of deficiency, so no case had yet been filed with the Tax Court. However, both GSK and the IRS wanted to obtain depositions from two key former GSK executives—Sir Paul Girolami and Sir David Jack—who had intimate knowledge of the company’s financial and R&D workings over several decades. If one person could be credited for inventing a particular drug, Jack was the inventor of Zantac. The Tax Court approved the joint motion based on the two executives’ advanced ages and the fact that their testimony was vital to a case that was likely to be filed. Although it was not specifically mentioned in the tax case, the documentation trail for the development and ownership of Zantac was apparently lacking, making the testimony of the two executives invaluable (“U.S. Marketing Intangibles Stance in Glaxo: Wave of the Future?” Tax Management Transfer Pricing Report, Vol. 12, No. 21, March 17, 2004).

In 2004, GSK sued the IRS, claiming it had erred in increasing GSK’s income by $7.8 billion ($4.5 billion for cost of goods sold, $1.9 billion for royalties, and $1.4 billion for interest income) for intercompany transactions involving Zantac, Ventolin, Ceftin, Zonfran, Imitrex, and Serevent. GSK later added a claim for a $1 billion refund based on a charge of discriminatory practices in the APA process. This charge stemmed from the SKB APA for Tagamat, to which GSK had access after the 2000 merger. The IRS was criticized for handling inbound and outbound transfer pricing differently.


By September 2006, the case was finally scheduled to go to trial in February 2007 after a delay based on claims of GSK withholding documents and a flood that destroyed some of the litigation information stored on IRS servers. On Sept. 11, 2006, the IRS and GSK announced a settlement concerning the transfer-pricing dispute for the tax years 1989 to 2000 and an agreement concerning the tax years 2001 to 2005. Under the settlement, GSK will pay the IRS $3.4 billion (or 60% of the total amount contested) and will drop its claim for a $1.8 billion refund. This represents the largest settlement in
IRS history (News Release IR-2006-142, 09/11/2006).

Lessons Learned

The IRS is serious about transfer pricing. Then–IRS Commissioner Mark W. Everson made the following statement concerning the GSK settlement: “We have consistently said that transfer pricing is one of the most significant challenges for us in the area of corporate tax administration. The settlement of this case is an important development and sends a strong message of our resolve to continue to deal with this issue going forward.”

IRS Chief Counsel Donald Korb reiterated the IRS’s commitment to settling transfer pricing disputes and said this about the settlement: “Our decision to accept GSK’s settlement offer reflects our commitment to resolving transfer pricing controversies without litigation, provided that our ultimate goal of compliance is not compromised.” The lawsuit’s 14-year duration backs up both statements.

Documentation is just as important for intangible assets as tangible ones. GSK might have been able to support its argument concerning the value of R&D if it had been able to prove how much had actually been spent in R&D. In particular, keeping track of the costs in the pharmaceutical R&D can be challenging. Few drugs are researched or developed in isolation. Not only are other drugs involved, but over the years many different entities may also be involved (for instance, different researchers, laboratories, and acquired projects). On the other hand, it is also difficult to document the value added by a company’s marketing efforts. Market studies and other documented efforts of a company’s marketing department should be helpful.

The APA process usually works, but it is not perfect. In 2006, 82 APAs were executed, of which 40 were bilateral agreements with other tax treaty countries. Revenue Procedure 2004-4 favors bilateral APAs; it requires that a request for a unilateral APA covering transactions with a related entity in a jurisdiction with which the United States has an income tax treaty must include “an explanation of why the request is not bilateral.” The GSK experience is an example of an unsuccessful multijurisdictional APA application.

Sharon Burnett, PhD, CPA, is a lecturer in accounting at Oklahoma State University, Stillwater, Okla.
Darlene Pulliam, PhD, CPA, is the McCray Professor of Accounting at West Texas A&M University, Canyon, Texas.




















The CPA Journal is broadly recognized as an outstanding, technical-refereed publication aimed at public practitioners, management, educators, and other accounting professionals. It is edited by CPAs for CPAs. Our goal is to provide CPAs and other accounting professionals with the information and news to enable them to be successful accountants, managers, and executives in today's practice environments.

©2009 The New York State Society of CPAs. Legal Notices