COMPARABLE PROFIT INTERVAL METHOD DOES NOT WEAR WELL WITH COSMETIC INDUSTRY.

Kavanaugh also states that the CPI is an IRS audit tool, not a method of determining intercompany pricing. Taxpayers, he says, need to know if their intercompany prices are reasonable at the time they are set. Kavanaugh asserts that, at best, CPI should be an optional method.

According to Kavanaugh, the matching transaction method (MTM) will also not work in his industry. Cosmetic companies go to great lengths to ensure the uniqueness of their products, he says, and, thus, the intangible being measured in a MTM analysis will always be different from that of any other competitor. Kavanaugh states that it is therefore critical to his industry that the comparable adjustable transaction method (CAT) be made more workable, and he says that this can be done by eliminating the requirement of a separate CPI verification of the results under CAT. Kavanaugh also writes that the regs' use of CPI as a cornerstone will cause severe problems to his industry with respect to tangibles.

We hereby submit the following comments on the recently released Section 482 proposed regulations, on behalf of the Cosmetic, Toiletry, and Fragrance Association ("CTFA").

CTFA is the national trade association representing the cosmetic, toiletry and fragrance industry, a $20 billion industry in the U.S. Our members include both U.S. based companies and foreign based companies with U.S. operations. In view of the global nature of the industry, intercompany pricing is of keen interest and importance to our members. These regulations will have significant impact upon our industry, as well as many other industries.

It is the strong recommendation of CTFA that the proposed regulations should be withdrawn because:

A) they exceed Congressional intent in that they apply to all intangibles and are not restricted to "high profit intangibles," and

B) they are not workable in the context of our industry.

Our industry is one that is characterized by:

o Primary reliance on marketing intangibles that do not generate "super" or extraordinary levels of profits.

o Cross border, related party transactions between entities EACH of which has developed or acquired, and enhanced, significant intangible assets.

o Product launches at widely varying and irregular intervals that are extremely costly, the effect of which produces uneven financial results.

o Substantial differences in strategies, outlets and targeted markets.

o A significant number of companies that are privately owned and do not make their data public.

o A significant number of companies that are engaged in a multiplicity of business lines, only one segment of which is in our industry, and whose published information generally integrates the data directly related to our industry's business lines with data not so related.

A) The proposed regulations exceed Congressional intent in that they apply to all intangibles and are not restricted to "high profit intangibles."

The proposed regulations go beyond Congressional intent by changing the rules not only for high profit intangibles, but for all intangibles (whether high profit or not), as well as for most transfers of tangible property. The scope of the regulations should be the same as the scope of the underlying legislation, i.e., high profit intangibles.

The proposed regulations are intended to implement the following amendment to Section 482: "In the case of any transfer (or license) of intangible property . . ., the income with respect to such transfer or license shall be commensurate with the income attributable to the intangible." Legislative history is clear that the amendment was targeted for "high profit intangibles." (See "General Explanation of the Tax Reform Act of 1986" prepared by the Staff of the Joint Committee on Taxation ("Blue Book"), pp. 1013- 1017, copy enclosed.) Congress perceived abuses in certain intercompany transfers of high profit intangibles and subsequent payments by the transferor to the transferee for the transferred intangibles; Congress then aimed to eliminate such abuses by inserting into the statute the "commensurate with the income attributable to the intangible" amendment.

We understand the difficulties that were presented to the drafters of the proposed regulations in limiting their scope to high profit intangibles since this category has never really been defined. However, although defining what is a high profit intangible, within the context of the legislative history and case law, is not so difficult. Using our industry as an example, high profit intangibles do not exist in the industry since:

o the sheer level of the licensor's expenditures are significantly lower than those incurred in industries with high profit intangibles, and

o the licensee expends significant funds on its own to enhance the value of the intangible.

Key intangibles involved in our products generally are marketing intangibles which are not high profit intangibles.

We believe, and Congress intended, that any regulations that try to determine the appropriate intercompany transfer pricing of high profit intangibles should not subject intangibles which are not high profit intangibles to arduous economic analysis and recordkeeping. As currently written, the proposed regulations apply to all intangibles. The statutory amendment was geared to a perceived abuse, the underlying regulations should be similarly aimed. Accordingly, these regulations should be withdrawn and rewritten to define the area of abuse as it relates to high profit intangibles.

In the case of an intangible that is not a high profit intangible, the existing regulations adequately address the determination of an adequate arm's length transfer price. This point was recognized by Congress.

B) The proposed regulations are not workable in the context of our industry.

The proposed regulations should be withdrawn and rewritten in a number of areas to make them workable for members of our industry. Changes that need to be effected include:

1) The Comparable Profit Interval Method ("CPI") must be overhauled.

a) CPI establishes a "commensurate with the income attributable to competitors" standard, rather than the mandated "commensurate with the income attributable to the intangible" standard.

b) CPI, by its very nature, is not a pricing method, but an IRS audit tool.

c) At best, CPI should be an optional method.

d) CPI tests should be calculated on an aggregate basis.

2) A profit split analysis must be allowed as an alternative method.

3) Other pricing methods described in the proposed regulations must be modified.

4) Qualified Cost Sharing rules must be made more flexible.

1) CPI must be overhauled.

a) CPI establishes a "commensurate with the income attributable to competitors" standard, rather than the mandated "commensurate with the income attributable to the intangible" standard.

The "commensurate with the income attributable to the intangible" standard means that a taxpayer must insure that the income it earns bears a reasonable relationship to the total income earned in connection with a particular intangible being transferred. In contrast, the proposed regulations ignore the income attributable to a particular intangible and, instead, require the taxpayer to set prices based on the income of competitors and their intangibles.

The CPI concept assumes that comparably situated companies should (in the absence of transfer pricing) earn similar profits. Numerous factors other than transfer price affect companies' margins in our industry. Different factors will apply to different companies, and the extent to which any one factor affects a company will differ from one entity to another. Such factors include, but are not limited to:

o Market share and placement

o Market acceptance of product

o Marketing and promotional techniques

o Effectiveness of marketing staff, advertising campaigns, etc.

o Quality of product

o Granting, acquisition, or expiration of a patent

o Sales volume

o Emphasis on margins vs. sales

o Reduction of prices to enter a market or to boost market share

o Price wars or price cutting by competitors

o Inventory management and inventory turnover

o Regional differences in operating costs

o Manufacturing efficiencies or inefficiencies

o Cost containment programs

o Financing costs

o Income from excess cash and other investments

o Income or loss from other unrelated business investments

o Business risks

o Contractual obligations

o Unanticipated events

o Recession or inflation

o Exchange rate fluctuations

o Efficiencies of management

o Management style

o Management philosophies/goals

o Short-term vs. long-term orientation

Differences in operating results due to market acceptance are particularly important in the case of companies dealing with branded name products, as is the case in our industry. Even a multiple year average does not provide a fair comparison since in our industry any one company could have minimal success (or failure) with its products over the period while another company had all successes during that same period.

The use of the CPI concept invariably will result in IRS adjustments whenever a company suffers a period of losses or reduced profitability. Such reduced profitability could be due to a declining market acceptance of products, failed product launches, poor management control of expenses, or any number of other business factors, none of which are even remotely related to transfer pricing. In fact, CPI in most cases will result in an adjustment whenever a taxpayer has a loss year despite multiple year averaging (unless the loss is slight and the other years are higher than normal to pull up the average). The CPI concept will force U.S. companies to show profits even when the total group shows a loss from the products sold into or out of the U.S. In addition, this adjustment would occur despite the fact that a company could have a loss in a year of a major launch.

The costs and timing of product launches will most dramatically impact comparability between companies. Our industry is characterized by multi-million dollar product launches wherein expenditures are made within a short time frame. The success of a new product cannot be guaranteed. It has been estimated that, with respect to fragrances, for example, the opportunity for success has decreased to one in five launches in recent years, compared to three out of five launches only five years ago. There is no set pattern for product launches. Different companies launch products at different intervals. The existence of a product launch and the rate of success or failure of launches will greatly impact a company's financial results.

A company with strong ongoing sales of existing products will be less impacted by new product launches than a company experiencing declining sales of existing products. Small companies with limited product lines will show far greater impact from new product launches than large companies with diverse product lines.

Accordingly, because of the nature of the business, companies in our industry will not be comparable, no matter what the time period is for comparison.

b) CPI, by its very nature, is not a pricing method, but an IRS audit tool.

The proposed regulations, which rely heavily on the CPI concept, purport to give guidance to a taxpayer in setting its intercompany pricing. CPI, however, is in fact nothing more than an IRS audit tool, and not a method of determining intercompany pricing. In the business world, prices are set before the beginning of the year, at a time when the final results of the prior year are not yet known. Thus, prices to be used in the target year must be set in the prior year at a time before the prior year's results are known. The proposed regulations require that intercompany prices for the target year are to be based on detailed CPI computations for the prior year, the target year and the year after the target year (the last two years not even having started yet). CPI computations are not available for any of the three years at the time the taxpayer must set its intercompany prices. These computations, even if in an industry for which CPI data can be retrieved readily, will not be able to be determined until several years after the transaction has taken place. At that point, due to the risk of high penalties, the taxpayer will be required to file amended returns, most likely without the benefit of correlative adjustments in the correspondingly affected foreign jurisdictions. If the audit of the year in question is current, then a timely amended return may not even be possible, thereby precluding the avoidance of penalties.

A taxpayer needs to know whether its intercompany prices are reasonable at the time they are set. As the regulations are currently drafted, a taxpayer will have to guess what the CPI data will be (perhaps by extrapolating from the data of comparable companies, if any can be found for the three years prior to the year in which prices are being set). If the market changes, or if an unforeseen event happens that impacts the comparable company data, then the taxpayer may find that its intercompany pricing, which was reasonable when set is now subject to challenge ex post facto. This can easily happen when the data relating to the years in question is finally in.

Just as in the case with third-party transactions, a taxpayer should be able to set its intercompany prices utilizing the most pertinent data available at the time of the transaction. Otherwise, the risk of double taxation is significantly increased because, once prices are set, they will be very difficult to change since retroactive change is a concept that is not accepted in arm's length transactions between third parties.

c) At best, CPI should be an optional method.

The CPI concept will not produce accurate results when applied to many companies in our industry. Therefore, CPI should be an optional method.

In order to find comparable companies in our industry, one would have to find entities in the same country/region that:

o have the same depth of product lines as the tested party,

o have launched a similar number of products during the period, and

o have similar levels of sales and market positions for existing products.

Consequently, it is highly unlikely that any tested company in our industry could find comparable companies.

In determining the applicable business classification to compute CPI, the regulations discuss determining whether there is "insufficient reliable data." However, no mention is made in the regulations as to whether such data must be relevant to the tested party. Thus, if it is necessary to broaden the applicable business classification to obtain "sufficient reliable data," as the proposed regulations require, a company in our industry will most likely be tested using data that is not at all relevant to its operations and, therefore, not comparable.

In our industry, broadening the business classification to obtain enough data will result in comparisons with companies having such disparate products and markets as to render the comparison meaningless. For example, SIC 2844 (Manufacturer of Cosmetics, Perfumes, Toiletries) includes products from shampoos purchased at a discount job lot store to high-end fragrances sold only at select upscale department stores or boutiques, as well as certain pharmaceutical products. SIC 5122 (Wholesale/Distributor of Drugs (includes Cosmetics)) covers cosmetics as well as bandages, vitamins and toothbrushes.

The lack of publicly available specific line of business data in our industry is a major impediment to application of CPI. Many of the major companies in our industry are owned by large conglomerates (pharmaceutical or consumer goods concerns) or are private concerns. Much of the publicly available information is not segmented along product or divisional lines and information of privately held companies will generally not be available in any public data base. For example, Exhibit A lists companies that were found in a data bank search for SIC 2844 and SIC 5122. It is clear from the listing that many companies that are within these SIC categories are not in our industry, and therefore are prima facie not comparable. A further analysis of the remaining companies shows that most engage in a number of business lines and their published financial results show commingled data of all their business lines (see Exhibit B). This is a dramatic illustration of the unavailability of comparable companies that can be used to construct a CPI in our industry.

Another major problem that must be addressed is the fact that, companies in our industry have significant intercompany transactions which cannot be used for comparability purposes. While intercompany transactions are eliminated in producing consolidated financial statements, the published data is nevertheless meaningless because it will include varying levels of sales (e.g., distributor, wholesaler, retailer and consumer). Further, even if separate company data were available, it would include so many intercompany transactions that it would also be meaningless.

Based on the foregoing, CPI information will generally not result in valid comparables, or will generally be totally unusable. However, taxpayers should be able to use CPI as an optional intangible and tangible pricing method which would be on an equal footing with all other methods in the regulations, in the absence of a Comparable Uncontrolled Price Method ("CUP") or Matching Transaction Method ("MTM"). CPI may be usable in some industries where there are a large number of companies that produce goods that are essentially homogeneous.

CPI is an attempt to apply one mechanical method of testing transfer pricing to a broad spectrum of taxpayers. At issue is whether any one method of transfer pricing will produce an equitable result for all taxpayers. Disparities between entities, industries, markets, etc., lead to the clear conclusion that no one method can possible satisfy all situations. The White Paper declares "(i)n the highly factual section 482 context, no one safe harbor or combination of safe harbors has yet been proposed that would be useful but not potentially abusive." White Paper, p. 78. This conclusion applies equally to any one mechanical method such as CPI to which a majority of taxpayers would become subject.

Although CPI is a mechanical test, it is far from objective. It will require use of subjective analysis of the data of both a taxpayer and comparable companies. It will require subjective decisions as to the appropriate comparable companies, the appropriate profit level indicators, the appropriate range, etc. Under CPI, disputes with the IRS will move from those based on a taxpayer's underlying fact situation to those involving the computation of CPI. The issues will involve extremely technical economic data, resulting in controversies between the taxpayer's and the IRS's economists, and very often will simply lead to the wrong answer.

CPI effectively results in disparate treatment between companies that have cross-border intercompany transactions and those that do not have such transactions. A company without cross-border intercompany transactions can certainly have its results fall outside of a comparable converging range, yet a company with such transactions cannot. There are reasons unrelated to intercompany pricing why companies may fall outside of the range. These reasons apply equally to companies with and without cross-border intercompany transactions. It is incorrect to assume that the reason why a tested party falls outside of a range is attributable to its transfer pricing. The listing on pages 6-7 specifically enumerates some of the multitude of possible causes for a deviation from the average. It is unfair and inequitable to require some companies to be taxed on profits that are determined based on such ranges when their competitors do not have a similar burden.

The White Paper indicates that "(t)he general goal of the commensurate with income standard is, therefore, to ensure that each party earns the income or return from the intangible that an unrelated party would earn in an arm's length transfer of the intangible." White Paper, p. 47. CPI analysis, however, is applied only to one party to a transaction. This, in effect, ignores the proper allocation of the overall profit or loss between the parties. The White Paper recognizes that "(l)ooking at the income related to the intangible and splitting it according to relative economic contributions is consistent with what unrelated parties do." White Paper, p. 47. It follows then that CPI would not be consistent with what unrelated parties do, particularly since unrelated parties do not set their prices based on the operating results of other companies.

For taxpayers who are fortunate enough to find sufficient data to construct a CPI range, and then have their transfer prices fall within a restricted range where the data converges, CPI may serve a useful function of providing assurance that the taxpayer has a comparable level of profitability as other similarly situated taxpayers. The use of CPI should, however, be expanded to allow for deviations. This can effectively be accomplished by considering CPI as a stand-alone method for intangibles and a distinct fourth method for tangibles. This concept is especially relevant to our industry since there are a limited number of competitors, limited access to data (i.e., privately held companies and companies controlled by conglomerates where industry specific data by brand or divisional lines cannot be obtained), and large variations in the mode of operations. For example, companies that sell to supermarket/drug chains may discount heavily, while companies that restrict distribution would not discount, and companies that sell door-to-door have a different expense structure than companies selling only to exclusive department stores.

The proposed regulations, while perhaps reducing some controversies, add new levels of controversies that previously were not relevant. A taxpayer may set a transfer price which falls within a certain CPI range, only to find that the IRS has constructed a different CPI and will look to a different point in its own CPI range. Allowing the taxpayer to choose or not to choose CPI depending upon its particular facts and circumstances, is the most appropriate assurance that the proper price has been set. CPI will not resolve disputes -- it will create them.

The current regulations specifically list factors which should be taken into account in arriving at an arm's length price, including uniqueness, value of services, availability of substitutes, prevailing rates, efficiencies, market share, etc. Profitability should be only one factor of many that is considered in determining an "arm's length" price.

The proposed regulations are unnecessarily restrictive by forcing taxpayers to use a CPI analysis in determining an appropriate transfer price (since MTM and CUP are virtually unattainable). This is particularly relevant to our industry, where many of the leading companies conduct business in materially different ways. In addition to the factors listed on pages 6-7 which will cause a firm to deviate from a CPI range, there are also large differences in philosophy in terms of new product launches, as discussed on page 8. There are even large differences between private and public companies, especially in terms of viewing profitability on a long-term or short-term basis. These factors affect a company's profitability compared to the industry and may, in fact, move that company out of the CPI range (e.g., a company may be willing to sacrifice short-term profits for long-term market share). Allowing for deviations from CPI provides for more accurate transfer pricing. The foregoing discussion applies equally to tangible property since in the proposed regulations tangibles have been swept up into the "commensurate with the income attributable to the intangible" standard.

d) CPI tests should be calculated on an aggregate basis.

Another major concern with the regulations is that they are written to test each related transferee on a separate basis. As a general rule, multinational companies establish their pricing on a worldwide basis. This is done for two primary reasons: administrative ease, and as an aid in justifying pricing in foreign jurisdictions. We firmly believe that any rules for testing the arm's length nature of intercompany pricing must recognize and accept the basic principle that is used today by most multinationals in establishing pricing -- the principle of aggregation.

Aggregation means simply that a taxpayer is permitted to establish a single intercompany price taking into account all intercompany affiliates on an aggregate basis. From a business perspective, most taxpayers do not and cannot take the ambiguities of each particular market into account in establishing its pricing. A taxpayer that sells to two international markets may be able to establish two separate prices to each of its affiliates accounting for various factors in each market, and could easily administer the system, but a multinational company with operations in, say, 30 to 50 locations (or more) could not possibly establish its pricing in the same manner. Instead, the multinational company aggregates the data from each market and establishes a single global price (or perhaps a few regional prices).

The uniformity which results from the aggregation principle is critically important to multinationals in defending their intercompany pricing arrangements to the foreign taxing authorities. The data generally initially requested by those tax authorities is a comparison of the prices charged to other related companies throughout the world. If the intercompany price (or royalty rate) is uniform, a presumption will generally arise that the pricing to the affiliate being tested is patently fair and, accordingly, arm's length. If, however, the pricing charged to the tested affiliate is different from, and in excess of, any other affiliate price, a presumption of invalidity arises and it is generally very difficult to overcome such a presumption. The ability to sustain pricing overseas will result in a reduction in Competent Authority disputes.

2) A profit split method must be allowed as an alternative method.

The proposed regulations effectively eliminate the possibility of using a profit split analysis and depart from the White Paper and numerous court cases which have utilized a profit split approach to apportion profits between related parties. The profit split method is discussed throughout the White Paper, and in fact, the White Paper correctly acknowledges that, in the absence of comparables, it is the most frequently used approach. Where both parties have developed or acquired, and enhanced, significant intangibles, the use of the profit split method should be expanded, not restricted.

The developer assister rules are a step in the right direction. We agree with the position taken therein that the contribution of each party to the development of certain intangibles, including the enhancement of trademarks and trade names, can be the subject of such rules. In our industry, this concept is particularly relevant because, although the parent company may develop and license its trademark or trade name, it is the subsidiary which enhances the value of that intangible in the local market. Example 4 in Prop. Treas. Reg. Sec. 1.482-2(d)(8)(iv) codifies this position and is particularly relevant to members in our industry since it illustrates the principle that the enhancer of an intangible, whether the enhancer is U.S. based or foreign based, is entitled to be compensated for such enhancement.

Cases such as Eli Lilly & Co., Inc., 84 T.C. 996 (1985), modified 856 F.2d 855 (1988), PPG Industries, Inc., 55 T.C. 928 (1970), Hospital Corporation of America, 81 T.C. 520 (1983), and G.D. Searle and Co., 88 T.C. 252 (1987) all stand for the proposition that profit split analysis is appropriate, and often is a preferred pricing method, when comparables are not available.

The White Paper specifically condoned use of the profit split method where both parties possess valuable intangibles, and provided a specific example of a French subsidiary of a U.S. corporation wherein pricing was tested utilizing a functional analysis and division of profits. See White Paper, pp. 99-102, (copy enclosed). In addition Appendix E of the White Paper (copy enclosed) provided three examples of instances where the profit split method is appropriate.

A profit split approach, by its very nature, requires both parties to be tested at once. A profit split analysis, as outlined in the White Paper more specifically, calls for an analysis of each function, risk and measurable assets of both parties. CPI, on the other hand, ignores the untested party and the relative value of functions performed by each party, and instead, compares the profit of the tested party to unrelated parties who may have relatively little in common with the tested party.

The profit split method has the benefit of taking into account the uniqueness of a particular company. This is particularly relevant in our industry where there are a limited number of companies, a shortage of publicly available data, and a great disparity among different companies and their markets.

In a non-abuse transaction (i.e., significant intangibles on both sides, neither of which is a high profit intangible) taxpayers should be able to rely upon the historically acceptable profit split method.

3) Other pricing methods described in the proposed regulations must be modified.

As a practical matter, the proposed regulations provide only two methods for the pricing of intangibles: MTM and CPI. This is notwithstanding the availability of a third method, the Comparable Adjustable Transaction Method ("CAT"). It is clear that this method cannot be considered a true stand-alone alternative because of the requirement that pricing resulting from a CAT must be verified using a CPI. As previously stated, CPI is not workable in our industry.

The only available intangible pricing method under the proposed regulations, therefore, would be MTM. However, given the conditions in the regulations which must be satisfied in order to qualify under MTM (which conditions are even more stringent than CUP found in the current regulations), it is highly unlikely that such method will ever be utilizable by members of our industry.

Our industry is characterized by, among other things, extremely strong product differentiation. Cosmetics companies go to great lengths to ensure the uniqueness of their products. Accordingly, the intangible being measured in a MTM analysis will always be different from that of any other competitor. The clear conclusion therefore is that the only circumstance in which MTM will ever be utilizable in our industry is where the same intangible is transferred under virtually identical conditions by the same taxpayer to both a related party (the tested transaction) and an unrelated party (the MTM transaction). We submit that this method will be virtually useless to our industry.

Accordingly, under the proposed regulations our industry is left with no realistic pricing alternative when testing intangible transfers -- MTM will be generally unavailable and CPI will be generally not workable.

CAT can and must be made more flexible so that it becomes a more viable method, usable in a greater cross-section of transactions. This can be accomplished by eliminating the requirement of separate CPI verification of the results obtained under CAT. A mechanical verification should not be required where the taxpayer has been able to identify a comparable third party uncontrolled transaction, albeit an inexact one.

TANGIBLES

The critical change made to the tangible pricing area is to require all pricing results, except for CUP's, to be verified by a CPI analysis. Given the serious problems which our industry has with CPI, the use of CPI as a cornerstone in the proposed regulations, adopted under the guise of consistency with the intangible pricing rules, creates an untenable position for us. We will have no viable pricing methods available to us in the absence of a CUP.

While the current regulatory framework is far from ideal in practice, it still provides taxpayers with workable parameters in which they can generally substantially fit their facts and circumstances. The proposed regulations will not provide viable alternatives to most taxpayers in our industry.

Therefore, the following modifications to the tangible pricing provisions of the regulations must be made:

i) CPI verification of pricing results determined under the cost plus, resale price, or any other fourth methods, must be eliminated.

ii) Except in cases where a CUP is available but not used by the taxpayer, the taxpayer's choice of any other pricing method should be presumed to be acceptable unless it can be demonstrated that such choice was unreasonable or that such method was inapplicable to the particular facts of the taxpayer.

iii) In cases where a tangible product has a material intangible content, the taxpayer should be given a choice of either using the intangible pricing methods or a bifurcated analysis, i.e., determine the transfer price for a generic product under one of the tangible pricing methods and then, where appropriate, make adjustments to account for the intangible component of a branded product.

4) Qualified cost sharing arrangement rules should be made more flexible.

"Qualified Cost Sharing Arrangements" (Prop. Treas. Reg. 1.482- 2(g)) ("QCSA's") may at first glance appear to be a welcome haven to taxpayers for avoiding the tangle of potential controversies inherent in the proposed regulations for transfer pricing of tangible and intangible property in general. Unfortunately, closer inspection reveals that the proposed regulations for QCSA's provide the IRS with a tool for seeking adjustments for deemed transfers of intangibles, rather than providing the taxpayer with a mechanism for avoiding transfer pricing controversies. Accordingly, to eliminate future controversies, the final QCSA regulations should be reworked, simplified, and made more flexible by providing a definitive set of guidelines which will give taxpayers and the IRS a high degree of certainty regarding the outcome of cost sharing arrangements.

NEED FOR FLEXIBILITY IN SCOPE OF AGREEMENTS

The proposed regulations on QCSA's are unnecessarily harsh and impractical in light of the realities of day-to-day business operations.

It should be within the discretion of the taxpayer to use a "blanket" cost sharing arrangement where appropriate. For example, it should be permissible to use language in the QCSA such as "pertaining to all products sold by the participant," when defining the marketing or manufacturing intangibles covered in a particular cost sharing arrangement. This flexibility is particularly important in our industry, where cost-shared intangibles are likely to be more numerous, varied, and less costly than, for example, in the automotive or computer industry.

On the other hand, taxpayers should be afforded enough flexibility to enter into QCSA's of a narrower scope (i.e., covering all intangibles developed for a particular product, or all products within a particular "brand"), where appropriate in light of the facts and circumstances of the business environment.

Providing taxpayers with a certain amount of flexibility with respect to defining the scope of QCSA's will NOT open the door for "cherry-picking," particularly if the "substantially disproportionate" test (discussed on pages 27-28) is retained in some form in the final regulations. On the other hand, even if the "substantially disproportionate" test is not retained in the final regulations, use of a "facts and circumstances" approach in defining the scope of a QCSA would not preclude the IRS from challenging the scope of an arrangement which is structured in a tax abusive manner.

"SPECIFIED INTEREST"

Another aspect of the proposed regulations which will prove unduly burdensome in practice is the requirement to transfer a "specified interest" -- or "enforceable legal interest" -- in any intangible developed pursuant to a QCSA. In many foreign jurisdictions, the transfer of a legally enforceable interest in an intangible may trigger a transfer tax, value added tax, or registration fee. Additionally, the transfer of an enforceable legal interest in an intangible may give a foreign taxing jurisdiction the mistaken impression that the expenditures in question give rise to a capital asset. It should suffice for the participant to benefit economically in an amount corresponding to the amount which would have been earned if the participant held a legal interest in the developed property.

"SUBSTANTIALLY DISPROPORTIONATE"

The proposed regulations do not define "grossly dispropor- tionate" (see Prop. Treas. Reg. Sec. 1.482-2(g)(2)(ii)(C)), and do not clearly explain the significance (if any) of the difference between a QCSA being determined to be "substantially disproportionate" (see Prop. Treas. Reg. Sec. 1.482-2(g)(4)(ii)(B)), or "grossly disproportionate."

It would appear from a reading of the proposed regulations that in the case of a "grossly disproportionate" QCSA, the entire arrangement is voided (which could result in a significant imputed royalty), while in the case of a "substantially disproportionate" QCSA, a "shared" royalty would be imputed. If that is the intention of the regulation, it should be clearly stated, since the result could be dramatically worse for a taxpayer under the "grossly disproportionate" test than under the "substantially disproportionate" test.

The "substantially disproportionate" test of the proposed regulations also discriminates against outbound taxpayers. Under the test, if a "U.S. participant" in a QCSA bears a "substantially disproportionate" (defined as where the U.S. participant's "cost/income" ratio is more than two times the "cost/income" ratio of the combined other participants) amount of the shared expenses, the IRS can deem that a transfer of an intangible has occurred outside the scope of the arrangement.

In addition, if the "substantially disproportionate" standard is retained in the final regulations, an appropriate mechanism must be devised to account for currency fluctuations among the countries included in a QCSA.

"U.S. PARTICIPANT"

In the case of U.S. based multinationals, the definition of "U.S. participant" (i.e., any controlled foreign corporation under Section 957), will likely cause every participant in a QCSA to be tested under the "substantially disproportionate" standard. It is not readily apparent WHY this definition of "U.S. participant" was chosen. Perhaps the drafters are concerned with the potential for outbound taxpayers to manipulate the amount of Subpart F income through cost sharing arrangements. However, it seems that in the case of an outbound taxpayer, shifting costs among controlled foreign corporation members of a QCSA will in most cases not result in any additional taxable income in the U.S., but only in adjustments to earnings and profits. For whatever reason the test was chosen, it is objectionable because it discriminates unfairly against outbound taxpayers, since most inbound taxpayers will have only one tested party. It would be much more practical, and equitable, for "U.S. participant" to be defined as "an entity which is subject to U.S. federal income taxes."

PARTNERSHIPS

Partnerships should be able to enter into QCSA's with other taxpayers. The partnership's share of expenses covered under such an arrangement should be shared by the partners in the same proportion as they share other income and expenses of the partnership.

If a partnership is established specifically for the purpose of developing an intangible, the partners should be considered "participants in a Qualified Cost Sharing Arrangement," if the arrangement otherwise meets the requirements under the regulations.

TRANSITIONAL RULE

The transitional rule, which allows taxpayers one year from publication to conform existing cost sharing agreements to the final regulations, does not provide an adequate amount of time. Multinational corporations (particularly "outbound taxpayers") will have to research the local tax, legal, and regulatory implications of amending cost sharing agreements in numerous jurisdictions. Agreements will likely have to be translated, reviewed by local counsel and tax advisors, and possibly registered and/or approved by local authorities. A longer period should be permitted.

CONCLUSION

We appreciate this opportunity to comment on the proposed regulations. Representatives of CTFA welcome the opportunity to meet with representatives of the Treasury Department to discuss our views.

Very truly yours,