Blog

A recent TaxNotes piece called for a substantial rewrite of Section 1.482-4, which addresses the transfer of intangible assets. The author, Ryan Finley, suggests that the Comparable Uncontrolled Transaction (CUT/CUP) approach should be relegated to a much more limited role. While many of the his assessments are fair, we would urge caution before relegating CUT too far to the backbench. There are certainly situations where CUT approaches are not only useful, but necessary as part of a larger framework to capture the issues and facts of the specific intercompany issue, examples of which we note later on.

European affiliates of multinationals such as IKEA face scrutiny from a variety of agencies including the European Union (EU), which issued EU Council Directive 2011/16 also known as DAC6. The stated purpose of DAC6, which became effective on June 25, 2018, is to provide transparency and fairness in taxation. DAC6 applies to cross-border tax arrangements between EU affiliates and tax havens. One of these cross-border tax arrangements is intercompany royalty payments from EU affiliates to affiliate in tax havens such as Liechtenstein. Such intercompany payments by European affiliates of IKEA are being challenged by the European Commission in a State Aid inquiry, which was initiated on December 18, 2017, according to an EC press release:

Like other expressions, economic categories reflect reality, and the term royalty isn't an exception.

Michiel de Vaa’s Etymological Dictionary of Latin and the Other Italic Languages (Brill, 2008) has no entry for royalty or its lemma. See his near-neighbor entry (Vaa, pp. 517-518): rego ("to direct, guide, govern"), but the listed cognates don't refer to payments or other economics allusions.

The Oxford English Dictionary, an arbiter of the English language, includes several meanings for royalty (noun (plural royalties)), and attributes its origin from Old French roialté, from roial (regal): 

“The sense ‘royal right (especially over minerals)’ (late 15th century) developed into the sense ‘payment made by a mineral producer to the site owner’ (mid-19th century), which was then transferred to payments for the use of patents, trademarks, and copyrighted materials.”

See the Oxford English Dictionary. Cite: “royalty, n.”

We find little empirical work in academia about the determinants of royalty rates covering intellectual property (IP) licensing transactions. Beyond the exchange of intellectual property itself, licensing transactions play an important role to market practitioners. Market practitioners benchmark their own pricing terms based on prior royalty rates transactions. Royalty rates are also important for those analyzing IP for transfer pricing and patent litigation, where comparable transactions are recognized by tax authorities and courts as benchmarks of value.

The claim that it is impossible to find comparable royalty rates and that from the start we should use the residual profit split method for high-value intangibles needs revisiting. This claim is made prima facie without regard for the license agreements available in curated databases such as RoyaltyStat, which as of today contains over 17,875 unique and unredacted license agreements. Also, adopting the residual profit split method, when separate tested party methods such as the TNMM could suffice, creates unwarranted costs and audit management challenges for both the taxpayer and the tax administration.

Naive analysts use the normal (Gauss-Laplace) distribution to compute the mean and standard deviation, or quartiles, without verifying if the data are abiding. Double fault is committed when they propose a “broad and unconvincing” range of data, such as the interquartile range (IQR), which may become useless to determine arm’s length (or reasonable) royalty rates. In practice, the interquartile range of royalty rates is used to help settle licensing disputes in tax or intellectual property valuation. 

The number of utility patents per license agreement primarily from two high-tech industries shows that most utility patent agreements include one patent number. The number of utility patents per agreement is a hyperbola. It's revealing that the interquartile range of royalty rates does not vary much considering the changing number of patents per agreement. In fact, the median royalty rate is stable at 5%, irrespective of the number of patents per agreement. This suggests that in addition to a stable median royalty rate,

Cross-country royalty rates based on the licensee's sales converge to a narrow range of median values from 6% to 10%, independent of sample size. The chart below shows the sample size (count) and median revenue-based royalty rates in several countries. The royalty rates on the chart convey several licensing rights, including patent, know-how, software, and trademark.

The two search criteria used in the RoyaltyStat database of unredacted license agreements were: (i) royalty rates must be based on net sales (revenue, turnover), and (ii) they must exclude related parties. Among the top eight countries, the U.S. has the biggest number of unredacted license agreements (count is 3,202, median royalty rate is 6%), followed by Canada (847, 7%), Mexico (384, 7%), China (339, 10%), Japan (256, 6.75%), U.K. (185, 9%), Germany (121, 9%), and France (108, 7.75%).

Overall, considering 13,025 unredacted license agreements with royalty rates based on net sales, excluding related parties, the median royalty rate is 5%; the lower quartile is 3.7%, and upper quartile is 10%. Data were retrieved from RoyaltyStat on March 17, 2016.

Roy Epstein & Paul Malherbe published a serious paper (2011) about comparable royalty rates used to determine patent damages in legal proceedings. However, they make several factual claims whose validity is compromised by the paucity of data examined. Despite the use of incomplete information, Epstein & Malherbe enumerate certain pitfalls of using royalty rates databases that we challenge as fallacious. See http://www.royepstein.com/Epstein%20&%20Malherbe%20-%202011.03.28%20-%20EPE%20(final%20AIPLA%20revisions).pdf

The center of gravity of royalty rates, measured by the median, is stable at 5% or 6% across industries. The sampled royalty rates are derived from unredacted license agreements from which the individual contractual terms can be compared. Except for Internet (Cyberspace) license agreements, a large sample (count = 13,004) of royalty rates shows that industry is not a key differentiator of the center of gravity. The cross-industry sample includes only royalty rates based on net sales, excluding related parties. Two hypotheses can be postulated subject to empirical testing based on resampling of the royalty rates data: