intellectual property 05 · 2020. 2. 1. · itself, makes to operating profit. this typically...

IntellectualProperty (4th Edition)

www.internationaltaxreview.com

Published in association with:

The Ballentine Barbera GroupErnst & YoungFTI ConsultingNERA Economic Consulting

Tax Reference Library No 24

www.internationaltaxreview.com22

Intellectual property: Royalties

Calculating royaltiesbased on comparablemarket opinionsBy Alexander Voegele,Stuart Harshbarger andNihan Mert-Beydilli,NERA EconomicConsulting

W hen a company licenses intellectual property (IP) to affiliate compa-nies, it is necessary to establish royalty rates. How does one calculatean arm’s-length royalty rate for intercompany licensing of IP? Thishighly complex issue can be approached from different viewpoints.

General business valuation of IP uses three principal methods: replacement cost,value in use, and fair market value. Replacement cost is related to how IP can be cre-ated, while value in use is based on what the ongoing business gets from investing inIP. Fair market value is based upon marketplace comparables and is the most fre-quently discussed and applied method. The OECD and most tax authorities requirethat companies adhere to the arm’s-length standard which can be demonstrated byuse of one or more of these three methods.

Replacement costReplacement costs are different from historical costs. Replacement cost is the amountof money that would be required to create similar or identical IP using resources thatare available today. Typically, this type of an approach will rely upon a total cost cal-culation that includes costs associated with unsuccessful attempts or dead-ends thatwere explored but not pursued. Replacement costs may be used in certain cases.Typically, practitioners restrict the use of this approach to very simple IP such as edi-torial rights and generic pharmaceutical products. Historical cost may be importantfor the determination of the economic ownership in the IP, but not for the determi-nation of its value.

Value in useApplication of this method is based upon calculating differentials between affectedor IP beneficiary products versus generic or non-IP affected products or services.The differentials identify a stream of incremental income that would not occurwithout the benefit of the IP. Once this stream of incremental value has been iden-tified and discounted, practitioners are able to calculate the IP’s net present valueor value in use.

The biggest drawback to using this approach to calculate intercompany royalties isthe lack of reliable financial data. In these instances, commercial value can be deter-mined only by a careful analysis of the incremental contribution that the asset, byitself, makes to operating profit. This typically requires comprehensive input fromfinancial personnel, marketing and sales departments.

Fair market valueFair market value refers to how much money the IP would fetch if it were sold orlicensed in a competitive market consisting of unrelated purchasers. For this approach

to produce reliable results, it must be the case that a viablecommercial market exists for the type of IP being valued.Within this market, there generally needs to be a large num-ber of firms who buy and sell this type of IP and informationregarding market participants and transacted prices should bereadily available and inexpensive. If these types of marketsare available for the IP being valued, then a fair market valueapproach can often be reliably applied. In practice, marketswith these types of open and competitive considerations onlyexist for the trading of durable goods not IP.

The transfer pricing methodsAll transfer pricing methods can be used to a certain extent tocalculate intercompany royalties. Practitioners will use thecomparable uncontrolled transaction (CUT) method, whencomparable royalties can be found, the cost plus (CP) methodfor common IP development and cost allocation systems, theresale price method (RPM) when IP is sublicensed to thirdparties, or a variety of the profit split method using a transac-tional net margin method (TNMM) approach when financialinformation is available for comparable companies. The prof-it split methods are the most widely applied in practice. Comparable uncontrolled transaction methodThe CUT method evaluates whether the amount charged fora controlled transfer of intangible property was arm’s lengthby reference to the amount charged in a comparable uncon-trolled transaction. To be deemed comparable under the CUTmethod, the intangibles involved in the controlled and uncon-trolled transactions must be used in connection with similarproducts or processes within the same general industry.

Internal and external comparables may be found in a fewindustries such as literature and music. Valuable comparableroyalties may even be found for pharmaceuticals from time totime since the quality of comparable data has recentlyimproved. For many types of IP, however, reliable data fromcomparable licensed IP is not yet in the public domain.

Comparable intangibles also must demonstrate similarprofit potential. Similar profit potential is most reliably meas-ured by calculating the net present value (NPV) of the bene-fits to be realized as a result of the transfer. ComparableNPVs can sometimes be found in joint ventures of pharma-ceutical companies.

The NPVs have to be calculated on forecasts using compa-rable data. Long-term forecasts and NPVs have to be changedfrom time to time since unforeseen economic phenomenaoften necessitate periodic alterations of forecasts. This is pre-cisely why IP agreements between unrelated parties will typ-ically incorporate stepped royalty regimes with detailedadjustment clauses.

To apply the CUT method reliably, the economic circum-stances and contractual parameters of the controlled anduncontrolled transactions must be closely comparable.However, this type of detailed information is often difficult

or impossible to obtain. As a result, the CUT method is most-ly used only as a back up or confirming method to one of theprofit split methods. In practice, some variation of a profitsplit method is almost always at the core of intercompanyroyalty rate calculations. Profit-split methodsComparable profit splitUnder the comparable profit split method, the shares of over-all or residual profit (or loss) are determined based on third-party evidence from an uncontrolled transaction in which thecombined return on assets is similar to the controlled situa-tion and there is a similar division of functions, costs, and risksbetween the independent parties in the uncontrolled situationand the affiliates in the controlled situation.

Comparable profit splits are often applied in the contextof royalty determinations as a way of determining the divisionof profit between licensor and licensee based on comparableindependent transactions involving similar products andintangibles. Examination of the pharmaceutical industryshows that there are some sectors of the economy with anabundance of third-party profit split transactions for use bytransfer pricing practitioners. If knowledge of these third-


www.internationaltaxreview.com 23

Alexander Voegele

NERA Economic ConsultingGrosse Bockenheimer Strasse 1360313 Frankfurt am MainGermanyTel: +49 69 133 8531Fax: +49 69 133 8535Email: [email protected]

Alexander Voegele is chairman of the advisory board of NERA EconomicConsulting.For over 25 years he has specialized in the development of innovativeeconomic structures for transfer pricing strategies, and in defending majorinternational transfer pricing cases.Voegele and his NERA colleagues provide economic transfer pricingstrategy and planning to clients facing M&As, dealing with organizationalstructuring and incentive management issues, and in need of market pric-ing evaluations. Having negotiated numerous bilateral and multilateralagreements involving Germany, France, the US, Japan, Canada, Mexico,and Australia, they have particular expertise structuring European arbitra-tion and advance pricing agreements. They specialize in worldwide docu-mentation, and valuing business and intellectual property.Voegele publishes articles and books on transfer pricing, including theleading German commentary, Handbook of Transfer Pricing. He speaks atconferences in Europe, the US and Asia.

Biography

party agreements is applied consistently to setting intercom-pany royalty rates, then arm’s-length results may be obtained.Contribution profit splitUnder this type of analysis, the combined profits are splitamong the participants in the value chain based on the rela-tive worth of the contributions associated with each partici-pant. To justify this split, the OECD Guidelines recommendthe use of external market data. In practice, the determina-tion of the relative value of each participant’s contribution caneither be based on external data or internal data.

Overall, this method is suitable when direct comparablescannot be used. When the split of the profits is based oninternal financial data to allocate profits, the method effec-tively results in “rate of return” pricing. Many tax authoritiesnow argue that such contributions should be calculated as thebuy-in payments to a cost allocation system based on animpartial investor model type of an approach.

Under the contribution profit split method, the combinedprofits (or losses) earned by two or more parties from a trans-action or series of transactions are divided based on some allo-cation principle; the resulting profit split is then implement-

ed by establishing the transfer price that accomplishes theindicated allocation. Diagram 1 illustrates such a calculationin the simple context of a profit split based on capitalemployed, where affiliate A grants IP to affiliate B.

As indicated in diagram 1, the affiliated group consisting ofaffiliates A and B has a consolidated operating profit of 60 onsales of 200 and capital employed of 300. Affiliate A has two-thirds of the consolidated capital employed. Thus, using acapital employed allocation, two-thirds of the consolidatedoperating profit is assigned to affiliate A and one-third isassigned to affiliate B. To produce this allocation, the interme-diate transfer price should be 120 (equal to affiliate A’s costof 80 plus operating profit 40 (= 2/3 x 60)).

Significantly, after the profit split, affiliates A and B will befound to earn the same rate of return on capital employed,which, in turn, will be equal to the overall return on capitalemployed earned by the affiliated group:

40/200 = 20/100 = 60/300 = 20%

The example illustrates the important general result that



Profit 40

Affiliate A

Profit 20

Affiliate B

Operatingprofit

60

Transfer price 120

Transfer price 200 Transfer price 200

Consolidatedvalue chain

Affiliate Bcapital(100)

Affiliate Acapital(200)

CapitalEmployed

Costs 80

20

40

Costs 80

Costs 60

Costs 60

Diagram 1: Contribution profit split method

under an asset-based profit split, the resulting rates of returnon the associated assets are equalized between the parties tothe profit split. In effect, this type of a profit-split computa-tion accomplishes a “rate-of-return” pricing result similar tothe comparable profits method, but with the consolidatedgroup’s overall rate of return, which is inherently arm’slength, serving as the benchmark for each affiliate’s arm’s-length rate of return in place of information for comparableindependent transactions.

Both the justification for using the profit split method inplace of alternative transfer pricing methods based on inde-pendent comparables and the specific approach taken inapplying the profit-split method must be grounded in eco-nomic logic, as well as the relevant facts and circumstances. Residual profit split methodAlternatively, the residual profit split method (RPM) uses atwo-step approach to determine an arm’s-length allocation ofoperating income or loss for both routine and non-routinecontributions. First, RPM requires an allocation of income to

provide a market return for routine contributions made by acontrolled taxpayer to the relevant business activity. Second,any residual profits attributable to non-routine intangiblesmust be allocated among the controlled taxpayers based onthe relative value of contributions of non-routine intangibleproperty to the overall enterprise.

Under the residual profit-split method, only a portion orresidual of consolidated operating profit—the portion emanat-ing from entrepreneurial or non-routine IP assets—is attributedto the affiliated parties based on profit split principles. Theremaining or routine amount of operating profit is attributed toroutine assets based on market benchmarks for functionally sim-ilar independent comparables. Diagram 2 illustrates the RPMconcept, assuming a return on capital employed (ROCE) forindependent functional comparables equal to 20% for affiliate Aand 12.5% for affiliate B. In parallel to the contribution profitsplit illustration in diagram 1, RPM equalizes the residual rate ofreturn on residual IP - capital across affiliates (equals 25% in dia-gram 2). “Residual IP-capital” here refers to investments in non-



Profit 30

Affiliate A

Profit 40

Affiliate B

Affiliate Beconomicinvestment

Affiliate Aeconomicinvestment

Entrepreneurial(residual)profit 30

ComparableROCE12.5%

ComparableROCE20.0%

Transfer price

Integrated value chain

Economicprofit split

Entrepreneurialcapital 160

Functionalcapital

100

Economic capitalemployed


Costs 70

20

10

Functionalprofit 20


Functionalprofit 20

Costs 70

Costs 60 Costs 60

ROCE = = = = = = 25%Residual profit

Residual capital

Residual profit

Residual capital

10 + 20

40 + 80

20

80

10

40

Diagram 2: Residual profit-split method concept

routine “entrepreneurial” IP-assets such as valuable IP that is notgenerally owned by uncontrolled taxpayers performing compa-rable functions to those of the controlled taxpayer.

The challengesMost studies will employ some type of CUT method as a firstpass toward final IP valuation. Sometimes it is possible to findcomparable royalty agreements in the public domain. In thesecases, the intangibles involved in the controlled and uncon-trolled transactions are used in connection with similar prod-ucts or processes within the same general industry.

Comparable intangibles also must demonstrate similarprofit potential. In practice, the profit potential of the com-parable intangibles is mostly unknown. Similar profit poten-tial is most reliably measured by calculating the NPV of thebenefits to be realized as a result of the transfer. If such com-parable NPVs can not be found, other methods have to beused in addition to the CUT method. In most cases, it is theRPM which is utilized in addition to a CUT analysis to calcu-late an arm’s length range of royalty rates.

Main weaknesses of using RPM to calculateintercompany royaltiesA main weakness already discussed is the potential for equal-ization of the residual rate of return to IP across affiliates as

shown in diagram 2. In addition, many economic studies thatuse RPM methods to calculate intercompany royalties assumethat the licensee only performs routine functions. These stud-ies also wrongly assume that licensees are only using one typeof IP in their operations.

No adjustments for non-routine functions of thelicenseeMost licensees of IP are complex entities. They do not onlyperform routine activities, but also non-routine activities, forexample, the typical pharmaceutical licensee normally per-forms several non-routine functions. The same applies tolicensees of brands. The presence of non-routine activities bythe licensees require adjustments which are rarely attemptedin practice.

No adequate adjustments for multiple types of IPat licenseeMany licensees of IP rely upon a diverse portfolio of IP toconduct their business in addition to what is purchased fromrelated or unrelated parties. Some of this IP may be devel-oped internally or licensed from other vendors. It is highlyunlikely that the licensee does not own other types of IP. Thisfact is sometimes ignored by practitioners or given insuffi-cient consideration.



Nihan Mert-Beydilli

NERA Economic Consulting875 North Michigan AvenueSuite 3650Chicago, IL 60611USTel: +1 312 573 2855Fax: +1 312 573 2810Email: [email protected]

Nihan Mert-Beydilli is a senior consultant at NERA, specializing in inter-company pricing and valuation analyses. Her areas of expertise includedesigning pricing methodologies, determining appropriate intercompanyprices (for example, arm's-length royalty rates), assisting clients to negoti-ate and implement APAs with tax authorities and supporting clients in taxcontroversy situations. She has advised multinational corporations in sec-tors including automotive, construction and industrial machinery, informa-tion technology services, and power generation. Mert-Beydilli has alsoworked on management consulting engagements and previously spentseveral years working as a qualified environmental engineer.Before joining NERA, Mert-Beydilli worked in the economics group ofAT Kearney's strategy practice. She holds an MBA from BrandeisUniversity and an MSc in environmental engineering.

Biography

Stuart Harshbarger

NERA Economic Consulting1157 Concord AvenueDetroit, MI 48207USTel: +1 313 267 9500Fax: +1 313 267 9501Email: [email protected]

Stuart Harshbarger is a vice-president in NERA's transfer pricing andintellectual property practices. He specializes in the economics associatedwith the manufacturing industries. He has completed global transfer pric-ing studies for US, European, and Japanese manufacturing companies,with an intensive focus on plant costing, performance, and profitabilitymetrics. Harshbarger has also completed transfer pricing studies for over30 different automotive parts manufacturers.Before joining NERA, Harshbarger worked for PricewaterhouseCoopers,where he was responsible for completing transfer pricing, intangible assetvaluation, and management fee studies for companies worldwide. He hasalso worked at DRI/McGraw-Hill, Argonne National Laboratory, theWashington Gas Light Company, and the US Department of Energy. Heis a sustaining member of the Detroit Economic Club.

Biography

Inability to account for changes in economicconditionsThe economy is a dynamic place that is continuously evolving.Typically, the amount of time necessary to complete an RPMstudy is so large that intervening changes have occurred in theunderlying economic assumptions making the final conclu-sions obsolete. And even if the NPV calculations are wellarticulated and robust, periodic adjustments to account forchanges in the marketplace are necessary to mimic thedynamic behaviour exhibited by competitive markets. Makingthese types of adjustments and keeping this type of complexdocumentation up to date can also be an expensive proposi-tion for companies

Inability to find economically reasonable systemFrequently, it is very difficult to find a convincing profit levelindicator or any other economic reasoning to determine anarm’s length royalty rate, for example, when splitting the netpresent value of a brand between licensor and licensee.

If, as is sometimes the case, application of pre-existingtransfer pricing methods such as RPM to a particular fact pat-tern do not work, then economic reasoning-type argumentswill generally be used by practitioners. Very often complicat-ed rules will be given that say in effect – according to theknowledge and experience of the analyst – a reasonable royal-ty rate for the IP being discussed is from 2% to 5% with amean of 3%. These types of discussions are usually done with-in the context of a formal presentation with a large amount ofsupporting documents and data, however, it is really just theopinion of the analysts involved in the project. A more objec-tive and hybrid arm’s-length approach would be to solicit theopinions of experts who are working in the particular indus-try being studied. One suggestion is that obtaining compara-ble market opinions from others is a more reliable methodthan accepting the opinions of a few consultants working onthe project.

Calculating intercompany royalties based oncomparable market opinionsThe benefits of consensus forecasting for macroeconomicvariables such as interest, unemployment, and inflation ratesare based on a simple premise: over the long run, the averageof many economic forecasts will generally be more accuratethan a single forecast. It is the benefits of consensus-basedforecasting that motivates the use of an expert witnessapproach to intercompany royalty rate calculations. The ideais not to have multiple consulting firms complete similar roy-alty studies and then use the average consensus of all the stud-ies to derive an arm’s-length intercompany royalty rate range.The approach is to identify expert observers both inside andoutside of a company who are qualified to offer an expertopinion on the amount of royalties an independent partywould be willing to pay for a particular type of IP. Ideally, out-

side expert observers should be employed by comparablecompanies within the same industry as the tested party. Theobjective is to find expert witnesses within the relevant mar-ket and to extract their opinions to obtain an arm’s-lengthrange of royalties.

Broadly speaking, this type of an approach with sufficientmodification, has been accepted by several European taxauthorities to find solutions during contentious tax audits andfor the calculation of arm’s-length royalties between relatedparties. This approach may also be used in the context of anadvance pricing agreement proceeding.

One of the largest difficulties of this type of an approachis obtaining data from unrelated third-party experts and pro-tecting the proprietary information of the tested party. Basedon experience, use of questionnaires, whether mailed or pre-sented in person, do not bear fruit. A more efficient approachis to interview outside and inside experts directly, accordingto a predetermined set of questions and information. Thoseinterviewed must have a detailed understanding of how thelocal markets work for the IP that the licensor is selling to thelicensee. The total number of interviewed persons should besufficient to be a statistically significant sample.

In many cases, economic studies of royalty rates runaground because too many comparable companies fail to dis-close certain information. When comparable data is missing,the economic criteria for the determination of such data haveto be carefully investigated. Typically, profit drivers and suc-cess factors have to be determined.

These economic criteria have to be translated into ques-tions. The number of questions has to be sufficient to obtainreliable results. An economic study of this type asked 20detailed questions to a group of 30 unrelated industryexperts.

Each question had to be answered with respect to itsinfluence on certain indicators, such as operating profits.The interviewed person was allowed to grant 1 point to 10points to each question. The interviewed person was askedto evaluate the importance of the individual question inrelation to other questions. He or she could decide that aquestion is, for example, three times more important thananother question. The total points for all questions leads toa maximum amount of achievable points, but each inter-viewed person determined the points assigned to eachindividual question.The total of all points achieved by allinterviewed persons leads automatically to a formula forcalculating a range of royalty rates. The challenge is toidentify the relevant profit level indicators, to ask the rightquestions and to identify a relevant group of inside andoutside experts.

This economic approach of obtaining comparable marketopinions was used to calculate the adjustments for non-rou-tine functions, to calculate the split of total IP into differenttypes of IP and to split IP between licensor and licensee.



Authorities’ consentIt is recommended that a comparable market opinionsmethod form a promising supplement to the application oftraditional valuation methods when there are informationaldeficiencies. It is advisable to call for the consent of govern-ing tax authorities to this type of an approach before the startof any studies. Several European tax authorities agreed to thisapproach when it was used to find solutions in tax audits aswell as for the calculation of future royalties.

The valuation of IP and the determination of royalties havedeveloped into one of the most contentious areas of transferpricing.

The comparable market opinions method described here issimilar to the use of economic experts within judicial pro-ceedings to assist judges in resolving complex IP tax litigation.Such a method may also provide, under certain circum-stances, a viable alternative to unilateral advance pricingagreements. The advantage of the comparable market opin-ions approach lies in the fact that it is based on the relevantexpert opinions resulting from potential transactions of com-parable companies.



For more information or to subscribe toInternational Tax Review please contact:

UK Hotline: +44 (0) 20 7779 8999 or US Hotline: +1 212 224 3570

intellectual property 05 · 2020. 2. 1. · itself, makes to operating profit. this typically...

Documents