A review of third-party license agreements: are periodic adjustments arm's length?

A Review of Third-Party License Agreements: Are Periodic Adjustments Arm's Length?

INTRODUCTION

On October 18, 1988, the Department of the Treasury and the Internal Revenue Service issued "A Study of Inter-company Pricing"(1) to explain the Government's position regarding the changes made by the Tax Reform Act of 1986 to section 482 of the Internal Revenue Code, pertaining to transfers and sales of intangible assets between related parties. The study responded to the congressional request, discussed in the Conference Committee report, for a review of the administration of section 482. The 1986 Act amended section 482 by adding the following sentence:

(1) U.S. Treasury Department (Office of International Tax Counsel, Office of Tax Analysis) and U.S. Internal Revenue Service (Office of Assistant Commisioner (International) and Office of Associate Chief Counsel (International), A Study of Intercompany Pricing (Discussion Draft, October 18, 1988)(hereinafter cited as "the White Paper').

In the case of any transfer (or license) of intangible

property (within the meaning of section 936(h)(3)(B)),

the income with respect to such transfer or license

shall be commensurate with the income attributable

to the intangible.

It also added a similar sentence to sections 367(d) and 936(h). These new provisions generally apply to taxable years beginning after December 31, 1986, but only with respect to transfers after November 16, 1985 (except for section 936(h) where all transfers of intangibles are covered regardless of the date of transfer).

The House Ways and Means Committee interpreted the commensurate-with-income language to include the necessity for "adjustments over time."(2) This interpretation is based on the belief that actual profit experience should be used in determining the appropriate compensation for the intangible and that periodic adjustments should be made to the compensation to reflect substantial changes in intangible income as well as changes in the economic activities performed and economic costs and risks borne by the related parties in exploiting the intangibles.(3)

(2) H.R. Rep. No. 426, 99the Cong., 1st Sess. 425 (1985).

(3) See White Paper at 48.

In addressing the congressional mandate, the IRS evinced the belief that contractual arrangements between unrelated parties contain some mechanism to change the terms of the contracts to take into account unanticipated events.(4) To support this position the IRS conducted a study of licensing agreements between unrelated parties. The IRS presented the results of its study in Appendix D of the White Paper.

(4) White Paper at 63. This position is contrary to legal precedent established in R.T. French Co. v. Commisioner, 60 T.C. 836 (1973). See Bausch & Lomb v. Commissioner, 92 T.C. No. 33 (Mar. 23, 1989), for additional Tax Court support of R.T. French.

The IRS conducted "[a] very preliminary review of unrelated party licensing agreements obtained from the files of the Securities and Exchange Commission [SEC] . . . [which] seems to support [adjustments over time]."(5) The IRS analyzed 60 of the 500 - plus agreements that had been filed with and were available at the SEC. The IRS explained in the White Paper that "[t]he choice of agreements in [its] sample was largely dictated by the ease of discovery and the availability of documents within the SEC's files." Two-thirds of the agreements the IRS examined are from Standard Industrial Classifications (SIC) for manufacturing firms producing "pharmaceutical preparations, toilet preparations, electronic computing equipment, semi-conductors, surgical and medical equipment, and ophthalmic goods." The other 20 agreements are for the services industry, mostly within the SICs for computer programming, computer software, and research and development laboratories.(6)

(5) White Paper at 63 (emphasis added).

(6) White Paper at Appendix D, 1.

Based on experience, many taxpayers and tax advisers doubted the validity of the IRS's conclusion that adjustments over time are arm's length. (Indeed, questions about the matter were fueled by the IRS's equivocal "seems to support" language.) In order to develop an empirical base from which to verify or challenge the White Paper's conclusions, a larger study (using a more representative sample) of licensing agreements was conducted. This article summarizes our analysis of the surveyed licensing agreements. In analyzing the agreements, we focused on the three issues relating to the adjustments-over-time concept:

* length of agreements

* termination and renegotiation of agreements

* compensation.

To support the IRS contention that licensors can force renegotiation of royalties when profitability of a licensed product unexpectedly rises, either (i) license agreements must contain termination or renegotiation clauses that can be activated by the licensor without showing cause or (ii) license agreements must be short-term agreements that can be renewed at their termination. The existence of neither of these conditions is supported by the data we analyzed.

Specifically, we found that agreements - particularly those involving patents and technology - are long term (using the IRS's definition of the term). In the vast majority of agreements between unrelated parties, the structure of royalty rates was known with certainty at the time the agreements were signed and could not be modified, except at the request of the licensee. The termination and renegotiation clauses contained in our sample rarely benefitted the licensor.

The next section of our paper discusses the IRS White Paper perspective concerning adjustments over time. Then we compare the results of our study with those of the IRS study. The article concludes with a summary and conclusion.

IRS WHITE PAPER PERSPECTIVE

Adjustments Over Time

In the White Paper, the IRS averred that adjustments over time are an integral part of arm's-length transactions between unrelated parties involving intangible property. More specifically, the IRS took the position that unrelated parties agree to renegotiate payments for intangible transfers as the profitability attributable to the intangibles changes. Therefore, in the IRS's view, a requirement for adjustments over time to related-party agreements is consistent with the arm's-length standard. On the basis of this position, the IRS proposed that taxpayers make adjustments to ensure there is a clear reflection of income between related parties.

We believe that adjustments over time are rare. For example, license agreements for patents or manufacturing know-how between unrelated parties typically require the licensee to devote research or manufacturing resources to the product that is licensed. In those instances, a licensee is unlikely to agree to an agreement that enables the licensor to cancel or renegotiate the agreement at his whim. To do so would put in question the return the licensee needs to receive in order to make the investment profitable for him.

At the same time, if the product under license is unprofitable for a significant period of time, the licensee may not be financially able to continue manufacturing the product after some point. Therefore, license agreements providing for cancellation without cause or renegotiation of the royalty rate would generally protect the licensee, not the licensor. In other words, third-party licenses could be expected to allow the licensee to force renegotiation of the royalty (to reduce it), but would rarely allow the licensor to force renegotiation merely because the product was more profitable than expected at the time of negotiation.

On this point, the risk of development of a market for a licensed product is not wholly that of the licensor. Typically, in third-party relationships, the licensor has the risk of development of the technology (risk during the R&D phase), whereas the licensee assumes the risk that the technology will effectively translate into a product that can be manufactured. In addition, the licensee bears the risk that the product will be acceptable to the marketplace.

The underlying premise of the adjustment-over-time concept is that licensors would desire to have some mechanism to adjust royalties as economic circumstances change. The IRS asserted in the White Paper that licensors would utilize this option to increase royalties when products become more profitable than they expected.

Renegotiation clauses and termination-without-cause clauses are the mechanisms that can be written into a license agreement that will force the adjustment over time which the IRS believes occurs. The following sections deal with each type of clause.

COMPARISON OF STUDIES

Methodology

We compiled and analyzed an extensive group of license agreements from SEC filings. The source of these agreements is identical to that cited by the IRS in generating its White Paper sample. We targeted industries on the basis of our expectations that these industries are typically involved in intercompany licensing of intangibles. The database currently includes 695 license agreements. Approximately 300 of these agreements were rejected because they involved related parties, were not applicable to the study, or were filed by both parties.(7) Thus, the sample is much larger and, we believe, more representative than that of the IRS.

(7) Agreements not applicable to the study included lease, consulting, and purchase agreements.

The first step in our analysis was to verify that each agreement in our sample was between unrelated parties. We also attempted to record the SIC code of each party in the agreement. We categorized agreements by SIC code to identify industries in which licensing is prevalent.

Our sample includes numerous agreements from several industries in which licensing appears commonplace. We categorized the 395 license agreements by the SIC code of the licensor or licensee. The licensor's SIC codes could be ascertained at the two-digit level for 252 agreements. The remaining 143 agreements are classified by the licensee's SIC code because it was not possible to determine a SIC code for the licensor. The sample includes agreements from 38 different industries. Seven industries account for 76 percent of the agreements. Table 1 lists the industries and the number of agreements. [Tabular Data Omitted]

Length of the Agreements

In the White Paper, the IRS attempted to make a case that adaptation to changing economic circumstances is implicit in many license agreements because of their short duration. The data the IRS cited in support of this position, however, do not show convincingly that the typical agreement is of short duration. In addition, its data are based on a small subset of the 60 agreements it analyzed.

Of the 33 agreements in the IRS sample containing termination-for-cause clauses, seven have specified lengths of less than three years and two are of five or six years' duration. Therefore, only 9 of 33 agreements allowing termination for cause are relatively short term (of six years' duration or less). This hardly supports the IRS's contention that such agreements tend to be of short duration. We can infer that 24 of the 33 agreements cannot be terminated except for cause within six years of the date they become effective, a finding that seemingly contradicts the IRS's contention that unrelated parties adjust the terms of their license agreements as economic circumstances change.

The IRS contended that unrelated-party license agreements are of "[a] relatively, short length of time." This, the IRS concluded, would allow automatic renegotiation if parties wish to extend the license. The IRS defined a short length of time as six years or less.(8) The data necessary to analyze this assertion were available from 189 agreements in our database.(9) Seventy-nine percent of these license agreements are in two categories: patent/technology or trademark/tradename.(10) We found these two categories had different characteristics with respect to length of agreement. Table 2 shows the distribution of these agreements. The 93 patent/technology agreements had an average term of 10.5 years. The distribution of these agreements reveals that 64 have a length greater than six years. Therefore, using the IRS's definition, 69 percent of the patent/technology agreements would not be considered an agreement with a short duration.

(8) White Paper at Appendix D, 9.

(9) The expiration date of the remaining agreements was tied to the life of the patent or was not stipulated. In either case the duration could not be calculated.

(10) The remaining agareements relate to software, marketing, distribution, or technical assistance.

Table : TABLE 2

               Length of Agreements
                     Number of Agreements
                    Patent/    Trademark/
Length in Years   Technology   Tradename
  1 - 3               12           37
  4 - 6               17            9
  7 - 9                6            2
  10 - 15             36            3
  16+                 22            6
  Total               93           57
  Average term      10.5          5.5

The trademark/tradename agreements had an average duration of 5.5 years. There are several potential explanations for patent/technology agreements' having longer lives than trademark/tradename agreements. It is conceivable that patent agreements must be of longer duration because of the investment in manufacturing capital that is required of the licensee in these situations. Thus, only with longer periods of time can the licensee recapture his investment and make an acceptable profit on that investment.

Termination

In the White Paper, the IRS endeavored to establish that license agreements typically include termination clause.(11) In its sample of SEC agreements, the IRS found that 20 (34 percent) of these agreements allow the licensee to terminate without cause and 13 (21 percent) extend this privilege to the licensor. Thirty-three agreements (55 percent of the sample) do not permit termination except for cause. In all, 50 of the 60 agreements in the sample have a termination clause of one type or another. (11) The IRS does not define termination for cause in the White Paper. We used termination for cause to mean that the license agreement can be terminated by one party because of the failure of the other party to adhere to certain aspects of the agreement, e.g., to remain a viable company, to manufacture and sell the product, etc. Termination without cause is the term used here to mean that one party or the other can terminate the agreement without giving a reason for the termination.

(11) The IRS does not define termination for cause in the White Paper. We used termination for cause to mean that the license agreement can be terminated by one party because of the failure of the other party to adhere to certain aspects of the agareement, e.g., to remain a viabel company, to manufacture and sell the product, etc. Termination without cause is the term used here to mean that one party or the other can terminate the agreement without giving a reason for the termination.

The IRS concluded that termination clauses support the adjustments-over-time concept because they offer the parties the opportunity to renegotiate agreements after they have been terminated. Termination of an agreement in order to renegotiate an increase in the royalty rate is most likely to occur where the licensor can terminate without cause. Therefore, we believe those agreements that offer the licensor the opportunity to terminate without cause are potential candidates for renegotiation.

The data in Table 3 suggest that approximately 85 percent of all license agreements contain some form of termination clause. Only four percent of the licensors, however, possess the ability to terminate an agreement without cause. This result differs sharply from that found by the IRS, which concluded that 22 percent of licensors could terminate without cause. Both studies support the conclusion that termination without cause is the exception. [Tabular Data Omitted]

Our data suggest that clauses permitting termination without cause are not widespread and that licensees are much more likely to have this option. Assuming that licensees would renegotiate only to lower royalties, the terms of unrelated-party agreements tend to favor reductions in royalties, rather than the increase required to support the IRS's White Paper position.

Renegotiation

One potential method to adjust licensing agreements is through renegotiation clauses. In reference to this topic the White Paper states:

"Although it is not possible to make general statements about the overall frequency of renegotiations or terminations based on the sample of agreements we examined, some examples of renegotiation were found."(12) (12) See White Paper at Appendix D, 10.

The IRS provided illustrative examples of this point but no statistical analysis. Thus, the White Paper mentions two agreements, one of which is cited below and the other which allows renegotiation under certain contingencies. The IRS noted that parties bound by an agreement may amend (renegotiate) that agreement even in the absence of a renegotiation clause and provides one example of this. These three examples of renegotiation, however, represent only five percent of the IRS's sample. The IRS reported only one example of a licensor's successfully renegotiating for a higher royalty during the term of an existing agreement.

Of the 395 agreements in our database, we found 13 agreements, or 3 percent, that contained renegotiation clauses. We examined each of these agreements to determine whether it favored the licensor or the licensee. The agreement favors the licensee if the renegotiation clause contains provisions that would lead to reductions in royalty rates. Seven agreements favored the licensee and the remaining agreements allowed for changes that might favor either party. The existence of such a small percentage of renegotiation clauses lends little support to the assertion that renegotiation provisions are a common occurrence in unrelated-party agreements. Furthermore, of the agreements providing this option over one-half benefitted the licensee.

The White Paper quotes several agreements to support its positions but does not cite the source of these agreements. One of these agreements that provides for renegotiation states:

"[B]oth the royalty rate and the scope of the Patented Portions will be renegotiated . . . [to] aggregate to not less than the 1.5 percent and not more than 2.5 percent of the Royalty Portion selling price of such Licensed products."(13) (13) This clause is from an agreement between Energy Conversion Devices, Inc. and Sony Corporation dated September 10, 1985.

(13) This clause is from an agareement between Energy Conversion Devices, Inc. and Sony Corporation dated September 10, 1985.

This agreement is one of the renegotiation agreements we identified as favoring the licensee. The reason for this classification is the fact that the original royalty rate stipulated in the agreement is three percent. In other words, this agreement does not support the IRS assertion that licensors renegotiate to increase the original royalty rate.

Compensation

There are three general types of compensation paid by licensees to licensors: (1) payments based on a percentage of sales revenues or units sold (royalties); (2) lump-sum payments; and (3) a variety of other forms of compensation. According to the statistics in Table 4, approximately 84 percent of the agreements provide for compensation in the form of royalties, 7 percent provide for lump-sums as the sole form of compensation, and 9 percent involve other forms of compensation.

Of the 60 agreements analyzed by the IRS, 46 (76 percent) stipulate that the licensee will pay a royalty to the licensor, 12 (20 percent) have lump-sum payments as the sole compensation, and 2 (3 percent) are reported to have other forms of compensation. Among the 46 agreements stipulating royalty payments, 21 establish constant royalty rates. The remaining 25 agreements have royalty rates that vary across time or over different sales levels. Seven of the IRS agreements specifying royalties based on net sales or units also require an initial lump-sum payment, whereas one-time or annual lump-sum payments are the only compensation provided for in 12 (20 percent) of the agreements analyzed by the IRS.

The IRS classified royalty payments as "constant" or "variable." Constant refers to the royalty rate being invariant with respect to both time and sales volume. The IRS reported 21 (35 percent) of the agreements in its sample have constant royalties. The IRS classified royalty payments as variable if rates vary over time or by sales volume. The examples of variable royalties include, ". . . a royalty rate that declines over time; others are structured so that the rate rises and then falls."(14) Twenty-five (41.7 percent) of the agreements in the IRS sample involve variable royalties. (14) See White Paper at Appendix D, 3.

In our view, alternative definitions of constant and variable royalties should be considered. The pertinent issue is not whether the terms of agreements change over time - they may change in a predetermined way - but whether they are altered in response to changing economic circumstances that were unanticipated at the time the license was negotiated. Therefore, we propose that agreements having a fixed royalty structure at the initial signing of the agreement be defined as having a "constant rate structure."

When two parties enter into one of these agreements they know with certainty the royalty rate at any sales level or at any point in time covered by the royalty. Those agreements for which the parties may alter the royalty terms during the period of time the agreement covers should be defined as having a "variable rate structure." To illustrate, if an agreement specifies a determinate, gradual reduction in the royalty rate over time, that royalty structure is classified as constant rate. Such an agreement does not lend support to the adjustments-over-time provisions of the White Paper.

Table 4 compares the IRS results with those in our study based on the IRS's definition of constant and variable royalties. It also presents the number of agreements in our database that have a constant rate structure. The results show that 93 percent of the agreements with royalties have a constant rate structure established at the initial date of the agreement. In other words, even though many of the sampled agreements have royalties that vary across either time or sales volume, the way in which these royalties are allowed to vary is determined at the initial date of the agreement. [Tabular Data Omitted]

SUMMARY AND CONCLUSION

We undertook our study to determine whether license agreements between unrelated parties support the adjustments-over-time concept in the IRS White Paper. We selected our sample from the same population used by the IRS. Because we used a larger sample size, however, we believe our results more closely reflect the actual characteristics of the population of license agreements. We found:

* In the majority of unrelated agreements, the structure of royalty rates was known with certainty at the time agreements were signed.

* The length of licensing agreements classified as patent and technology were not of short duration.

* Few agreements could be terminated without cause and when this option was available it was more likely to benefit the licensee, i.e., result in lower royalties.

* A very small percentage of agreements contained renegotiation clauses. More than one-half of these clauses favored the licensee; none clearly favored the licensor.

In each of the categories, we found little evidence supporting the adjustments-over-time concept and the assertions the IRS presented in the White Paper. We believe that the adjustments-over-time concept is not arm's length and should not be incorporated into the section 482 regulations.