Reproduced with permission from Tax Management In- ternational Journal, 48 TMIJ 256, 06/14/2019. Copyright 2019 by The Bureau of National Affairs, Inc. (800-372- 1033) http://www.bna.com Transfer Pricing Challenges in the Digital Economy: A Case Study of the Internet of Things (Part II of II) By Lorraine Eden, Niraja Srinivasan, and Srini Lalapet * Are the current OECD and IRS transfer pricing methods and guidance adequate for taxing the global profits of multinational enterprises (MNEs) in the digital economy? If not, how might U.S. transfer pric- ing practitioners in business, consulting, and govern- ment work together to modify the approaches to re- flect how value is created and measured in evolving digital business models? 1 In the digital economy, 2 value is not created in iso- lation by a company for the benefit of the customer but, in fact, is created as a consequence of the con- stant flow of information between the company and the customer. Value creation, therefore, is no longer a static eventuality at the end of a value chain but rather a result of dynamic interaction within a digital ecosys- tem of shops and networks. From a transfer pricing perspective, these new business models challenge transfer pricing and international tax practitioners to consider whether their existing frameworks still apply. This issue is critical as we move into a digital world characterized by continuous and circular data flows, value shifts, and greater functional complexity of re- lated and unrelated parties, both across space and time. This article is the second in a two-part series de- signed to illustrate the complexity of digital business models and the challenge of applying transfer pricing analyses based on the value creation approach. In the first article, we discussed the old and new firms in the digital economy and reviewed the OECD’s BEPS project focusing on Action Items 8–10 (value cre- ation) and 1 (the digital economy). 3 We argued that, given the newness and complexity of these new digi- tal business models, we need to better understand the challenges they create for applying a transfer pricing framework in the 2017 OECD Transfer Pricing Guidelines. 4 In this second article, to better understand the chal- lenges that digital business models pose for transfer pricing analysis, we examine a stylized case study drawn from the technology industry, specifically, from the emerging world of the Internet of Things (IoT). * Lorraine Eden is a Professor of Management at Texas A&M University ([email protected]). Niraja Srinivasan is Vice President of Global Tax/Transfer Pricing at Dell Technologies ([email protected]). Srini Lalapet is Director of Tax and Senior Transfer Pricing Economist at Dell Technologies ([email protected]). 1 We thank Reilly Smith for research assistance and William Byrnes for helpful comments. Early versions of this paper were presented at meetings of the American Bar Association—Tax Sec- tion, NABE Transfer Pricing Symposium, World Investment Fo- rum, and International Studies Association. The views and opin- ions expressed here belong solely to the authors, and not to their employers or to any other group or individual. 2 Eden, Lorraine, Multinationals and Foreign Investment Poli- cies in a Digital World, The E15 Initiative: Strengthening the Global Trade and Investment System for Sustainable Develop- ment, E15 Task Force on Investment Policy, World Economic Fo- rum and the International Centre for Trade and Sustainable Devel- opment (ICTSD) (www.e15initiative.org); Eden, Lorraine, The Fourth Industrial Revolution: Seven Lessons from the Past, Inter- national Business in the Information and Digital Age (Alain Ver- beke, Robert van Tulder & Lucia Piscitello eds.) (Progress in In- ternational Business Research, Vol. 13 (European Int’l Business Academy and Emerald Pub., 2019). 3 Eden, Srinivasan, and Lalapet, Transfer Pricing Challenges in the Digital Economy: Hic Sunt Dracones? (Part I of II), 48 Tax Mgmt. Int’l J. 251 (June 2019). 4 2017 OECD, Transfer Pricing Guidelines for Multinational Enterprises and Tax Administrations (Paris, July 2017). International Journal TM Tax Management International Journal 2019 Tax Management Inc., a subsidiary of The Bureau of National Affairs, Inc. 1 ISSN 0090-4600
Transcript
Reproduced with permission from Tax Management In-ternational Journal, 48 TMIJ 256, 06/14/2019. Copyright� 2019 by The Bureau of National Affairs, Inc. (800-372-1033) http://www.bna.com
Transfer Pricing Challengesin the Digital Economy: ACase Study of the Internet ofThings (Part II of II)By Lorraine Eden, Niraja Srinivasan, and Srini Lalapet*
Are the current OECD and IRS transfer pricingmethods and guidance adequate for taxing the globalprofits of multinational enterprises (MNEs) in thedigital economy? If not, how might U.S. transfer pric-ing practitioners in business, consulting, and govern-ment work together to modify the approaches to re-flect how value is created and measured in evolvingdigital business models?1
In the digital economy,2 value is not created in iso-lation by a company for the benefit of the customer
but, in fact, is created as a consequence of the con-stant flow of information between the company andthe customer. Value creation, therefore, is no longer astatic eventuality at the end of a value chain but rathera result of dynamic interaction within a digital ecosys-tem of shops and networks. From a transfer pricingperspective, these new business models challengetransfer pricing and international tax practitioners toconsider whether their existing frameworks still apply.This issue is critical as we move into a digital worldcharacterized by continuous and circular data flows,value shifts, and greater functional complexity of re-lated and unrelated parties, both across space andtime.
This article is the second in a two-part series de-signed to illustrate the complexity of digital businessmodels and the challenge of applying transfer pricinganalyses based on the value creation approach. In thefirst article, we discussed the old and new firms in thedigital economy and reviewed the OECD’s BEPSproject focusing on Action Items 8–10 (value cre-ation) and 1 (the digital economy).3 We argued that,given the newness and complexity of these new digi-tal business models, we need to better understand thechallenges they create for applying a transfer pricingframework in the 2017 OECD Transfer PricingGuidelines.4
In this second article, to better understand the chal-lenges that digital business models pose for transferpricing analysis, we examine a stylized case studydrawn from the technology industry, specifically, fromthe emerging world of the Internet of Things (IoT).
* Lorraine Eden is a Professor of Management at Texas A&MUniversity ([email protected]). Niraja Srinivasan is Vice Presidentof Global Tax/Transfer Pricing at Dell Technologies([email protected]). Srini Lalapet is Director of Taxand Senior Transfer Pricing Economist at Dell Technologies([email protected]).
1 We thank Reilly Smith for research assistance and WilliamByrnes for helpful comments. Early versions of this paper werepresented at meetings of the American Bar Association—Tax Sec-tion, NABE Transfer Pricing Symposium, World Investment Fo-rum, and International Studies Association. The views and opin-ions expressed here belong solely to the authors, and not to theiremployers or to any other group or individual.
2 Eden, Lorraine, Multinationals and Foreign Investment Poli-cies in a Digital World, The E15 Initiative: Strengthening theGlobal Trade and Investment System for Sustainable Develop-ment, E15 Task Force on Investment Policy, World Economic Fo-rum and the International Centre for Trade and Sustainable Devel-opment (ICTSD) (www.e15initiative.org); Eden, Lorraine, TheFourth Industrial Revolution: Seven Lessons from the Past, Inter-national Business in the Information and Digital Age (Alain Ver-
beke, Robert van Tulder & Lucia Piscitello eds.) (Progress in In-ternational Business Research, Vol. 13 (European Int’l BusinessAcademy and Emerald Pub., 2019).
3 Eden, Srinivasan, and Lalapet, Transfer Pricing Challenges inthe Digital Economy: Hic Sunt Dracones? (Part I of II), 48 TaxMgmt. Int’l J. 251 (June 2019).
4 2017 OECD, Transfer Pricing Guidelines for MultinationalEnterprises and Tax Administrations (Paris, July 2017).
International JournalTM
Tax Management International Journal
� 2019 Tax Management Inc., a subsidiary of The Bureau of National Affairs, Inc. 1ISSN 0090-4600
IoT has been defined as ‘‘sensors and actuators con-nected by networks to computing systems’’ where theconnected sensors and systems are used to ‘‘monitoror manage the health and actions of connected objectsand machines [and] the natural world, people, andanimals.’’5 IoT emerged from the convergence ofmultiple new technologies, including but not limitedto, embedded systems, artificial intelligence, real-timedata analytics, enhanced data virtualization, and in-creased storage capacity within cloud-based solu-tions.6 As such, IoT provides an excellent case studyfor us to study the challenges of transfer pricing in thedigital economy.
We provide an overview of the IoT system and out-line two new business models: the Direct BusinessModel and the Partner Business Model. We thenevaluate each model using a traditional transfer pric-ing analysis framework. We argue that IoT creates atleast four challenges for transfer pricing: data as anew type of related-party transaction; circularity ofand value shifts in the IoT data/insight exchange; thespeed of technological change and functionality; anddifficulty in characterizing control, decentralizationand cooperation among the related parties. We exploreeach challenge and conclude that more work isneeded to ‘‘lift the veil’’ on—let alone set up the rulesfor—transfer pricing in the digital economy.
THE INTERNET OF THINGS (IoT)The Internet is a ‘‘global system of interconnected
computer networks that use the standard Internet pro-tocol suite based on TCP/IP.’’7 IoT involves the appli-cation of the Internet to physical objects (e.g., sensors,vehicles, mobile phones, and home appliances) suchthat the physical objects gain the ability to autono-mously sense and communicate with other objects onthe same network.8 Some commentators define theIoT as ‘‘an open and comprehensive network of intel-ligent objects that have the capacity to auto-organize,share information, data and resources, reacting and
acting in face of situations and changes in the envi-ronment.’’9
IoT relies on the connectivity of devices and sen-sors through wired or short-range wireless networkssuch as RFID tags, Bluetooth and Wi-Fi.10 IoT de-vices are connected, intelligent devices that can beclassified as resource rich (e.g., smart phones andwatches, personal computers) or resource constrained(e.g., sensors, light bulbs, and switches), dependingon whether they do or do not have the hardware andsoftware capabilities on their own to support theTCP/IP Protocol and communicate across the Internet.Resource-rich IoT devices use a ‘‘device-to-cloud’’ ar-chitecture, which may or may not be mediatedthrough an IPv4/IPv6 security gateway. IoT devicescan also be set up in a multi-layered architecture. Forexample, in a three-layer architecture a ‘‘network’’layer of sensors is used to collect data from a ‘‘per-ception’’ layer of physical objects; and the sensorstransmit the data via a communications network (thecloud) to a ‘‘computation’’ layer to process and ana-lyze the data.11
An IoT solution also requires various technologiesto be fully integrated into a single operational whole.Perhaps the best way to think of an IoT solution is anecosystem of connected hardware, software and ser-vices that work in tandem to solve a specific problem.This connectivity may be enabled by the Internet, butone can also conceive of closed, non-Internet-basedcommunication systems where IoT can be operation-alized. In broad terms, we can think of IoT as havingfour distinct areas within the overall ecosystem thatenable an IoT solution to work, which we illustrate inFigure 1 and outline below:
• Customer: The customer both produces data forthe IoT solution and consumes the end product ofthat data after it has been analyzed and trans-formed into valuable insights. In an IoT solution,the customer is no longer just a passive recipientof the value that is created (as in the traditionalvalue chain model) but is, arguably, an integralparticipant in the value creation process generatedby an open innovation platform.
• Edge: The Edge refers to the devices, sensors,and similar equipment at the customer location
5 McKinsey Global Institute, The Internet of Things: Mappingthe Value Beyond the Hype, at 17 (McKinsey & Company, June2015).
6 Frost & Sullivan, Bridging the Gap Between Operations andInformation Technology: Accelerate IoT Solution Developmentand Deployment with Telit (whitepaper) (2018) (www.frost.com).
7 Madakam, Somayya, Ramaswamy, R., and Tripathi Siddharth,Internet of Things (IoT): A Literature Review, J. of Computer andCommc’ns, at 164 (2015).
8 Ara, Tabassum, Shah, Pritam Gajkumar, and Prabhakar, M.,Internet of Things Architecture and Applications: A Survey, IndianJ. of Sci. and Tech., at 9(45) (DOI: 10.17485/ijst/2016/v9i45/106507); 2015 OECD, Addressing the Tax Challenges of the Digi-tal Economy, Action 1: 2015 Final Report, Ch. 6, OECD/G20BEPS Project (Paris).
9 Madakam, Somayya, Ramaswamy, R., and Tripathi Siddharth,Internet of Things (IoT): A Literature Review, J. of Computer andCommnc’ns, at 165 (2015).
10 2018 OECD, IoT Measurement and Applications, OECDDigital Economy Papers No. 271, at 247 (Paris, Oct. 2018).
11 Ara, Tabassum, Shah, Pritam Gajkumar, and Prabhakar, M.,Internet of Things Architecture and Applications: A Survey. IndianJ. of Sci. and Tech., at 9(45) (DOI: 10.17485/ijst/2016/v9i45/106507).
Tax Management International Journal2 � 2019 Tax Management Inc., a subsidiary of The Bureau of National Affairs, Inc.
ISSN 0090-4600
that are used to collect the data. There are two ge-neric types of devices in the Edge: embedded sys-tems (i.e., the physical appliances and sensorsused to collect individual data points) and gate-ways (i.e., specialized hardware used to collatedata and pass it on to a data storage solution forfurther processing and analytics). While the Edgeis usually seen as the entry-point of an IoT solu-tion, it can also be the point where complex dataanalytics are performed and value is created.
• Core: The Core refers to both on-premise and off-premise storage solutions and servers (includingdata centers) that help to store and organize thecollected data for further analytics. The Core mayalso include teams of highly skilled data scientistsand data threat software that work on securing theintegrity of the data transferred from the Edge tothe on and off-premise datacenters and the Cloud.The Distributed Core can be the repository of rawdata as well as the analyzed data and is at theheart of an IoT solution exchanging data withboth the Edge and the Cloud in a dynamic andcontinuous value creation loop.
• Cloud: This refers to cloud-based applicationsand software used to analyze the stored data toprovide valuable insights. The Cloud can be con-ceptualized as the brain of the IoT solution wherespecific applications and software analyze thedata that is flowing from the Edge through to theCore and onto the Cloud. That said, IoT solutionscan also be designed such that data analyses areperformed at the Edge as well as the Core and theCloud as needed.
IoT offerings are exemplars of business models thatspan both the physical and the digital world and sit atthe nexus of the integration of operational technologyand information technology.12 Operational technology(OT) is ‘‘the hardware and software used in sensing
and collecting data. This includes all the hardware atthe edge of the network’’; whereas the informationtechnology (IT) in IoT includes ‘‘the network, cloud-based platforms, data analytics, and integration withother cloud-based platforms.’’13 Most IoT solutionsinvolve both IT and OT. An IoT solution would notexist without the constant interaction and exchange ofdata between the physical and digital worlds, and be-tween operational technology and information tech-nology. In fact, it is this interaction and data exchangethat are critical to the success of any IoT solution.14
The complexity of an IoT solution is further exac-erbated by interfirm collaboration—the opportunityfor multiple companies to participate in providing anintegrated solution to a customer. For instance, nu-merous technology companies participate in varyingdegrees in different parts of the IoT Ecosystem asnoted below (this is not an exhaustive list):
• Edge: Field gateways, sensors and appliance pro-viders (Dell, Emerson, Nest, GE, etc.);
• Core: Servers and storage solutions (Dell EMC,HP, etc.); and
• Cloud: Cloud software and services, security soft-ware, data analytics and applications (MicrosoftAzure, AWS, VMware, etc.).
The principal challenge is to integrate the variousparts of an IoT ecosystem into a coherent whole thatprovides value to the customer. To do this, an Indus-try 4.0 firm might adopt several business models, twoof which we examine here, the Direct Business Model(hereinafter, referred to as Direct Model) and the Part-ner Business Model (hereinafter, referred to as Part-ner Model).
• Direct Model—In this relatively simple model, asingle company sells sensors and Edge gatewaysdirectly to the end-customer and helps with theintegration of core and cloud offerings that arealso offered by the same company. In addition tothe collection of millions of specific data pointsfrom the customer, a key characteristic of mostIoT solutions is that other technologies such asmodern imaging, big data and predictive analyt-ics using machine learning may be used in tan-dem to process the data and make meaningful de-cisions.
12 2016 OECD, The Internet of Things: Seizing the Benefits andAddressing the Challenges, Ministerial Meeting on the Digital
Economy Background Report, OECD Digital Economy PapersNo. 252 (Paris); 2018 OECD, IoT Measurement and Applications,OECD Digital Economy Papers No. 271 (Paris, Oct. 2018).
13 Frost & Sullivan, Bridging the Gap Between Operations andInformation Technology: Accelerate IoT Solution Developmentand Deployment with Telit (whitepaper), at 3 (2018) (www.frost-.com).
14 Frost & Sullivan, Bridging the Gap Between Operations andInformation Technology: Accelerate IoT Solution Developmentand Deployment with Telit (whitepaper) (2018) (www.frost.com).
Tax Management International Journal
� 2019 Tax Management Inc., a subsidiary of The Bureau of National Affairs, Inc. 3ISSN 0090-4600
• Partner Model—In this model, which is com-monly used by many IoT solutions providers, acompany delivers an IoT solution by integratingthe infrastructure and hardware/software offeredby a partner or original equipment manufacturer(OEM) into their offerings to the end customer.Examples of this type of business model includemany industrial IoT solutions and ‘‘infrastructureas a service’’ (IaaS) solutions where multiplecompanies can be involved in providing an IoTsolution to a customer. In these business models,there is typically a convergence and collaborationof operational technology at the customer sitewith information technology within the IoT solu-tion.
Given the complexity of IoT business models,whether a Direct or Partner Model, and the fact thatIoT can be applied in a wide variety of consumer andenterprise applications, identifying how and wherevalue is created within an IoT ecosystem, and per-forming a transfer pricing analysis poses numerouschallenges. To illustrate these challenges, we examinea stylized industrial IoT case study (adapted to illus-trate both models) and evaluate it from a traditionaltransfer pricing analysis framework.
TRANSFER PRICING IN AN IoTECOSYSTEMTransfer Pricing Analysis
A transfer pricing analysis typically follows a se-quence of analytical steps. Performing these steps isnot easy and many controversies and disagreementscan arise between MNEs and tax authorities over the‘‘right’’ price or margin as a result of the transfer pric-ing analysis.15 for discussions of some of the difficul-ties.) The OECD Transfer Pricing Guidelines16 out-line the steps necessary to apply the value creation ap-proach to the arm’s-length standard as the following:
• Identification of the intercompany transactionsand the related parties involved in such transac-tions;
• An analysis of the functions performed, risks as-sumed, and assets employed (FAR analysis) in the
context of the identified intercompany transac-tions that lead to a characterization of the relatedparties as distributors, manufacturers, intangibleproperty (IP) owners, service providers, etc. Thiscould be supplemented by a value chain analysisor a DEMPE17 (development, enhancement,maintenance, protection and exploitation of intan-gibles) analysis which serves to identify whereand how value is created and by whom; and
• Performing an economic analysis including re-viewing and selecting an appropriate transferpricing method(s), followed by their applicationto determine an arm’s-length price or profit.
We now apply this framework to the Direct andPartner IoT business models.
Transfer Pricing in the Direct ModelThe Direct ModelIn its simplest version, the Direct IoT Model can be
a single company offering different facets of the IoTsolution through multiple legal entities located in vari-ous jurisdictions. A stylized example of the DirectModel is illustrated in Diagram 2 below.
In the above example, our case study assumes thatthe end customer has a multi-year business contractwith the IoT firm. Within the IoT firm are severalwholly owned legal entities that engage in intercom-pany transactions that result in providing the end cus-tomer with the IoT solution. In this illustration, Enti-
15 Eden, Lorraine, Taxing Multinationals: Transfer Pricing andCorporate Income Taxation in North America (Univ. of TorontoPress, 1998); Eden, Lorraine, The Economics of Transfer Pricing:The International Library of Critical Writings in Economics, Ed-ward Elgar Pub. (Cheltenham, U.K., 2019).
Eden, Lorraine, The Arm’s Length Standard: Making It Workin a 21st Century World of Multinationals and Nation States(Thomas Pogge and Krishen Mehta eds.), Global Tax Fairness(Oxford Univ. Press, 2016).
16 2017 OECD, Transfer Pricing Guidelines for MultinationalEnterprises and Tax Administrations (Paris, July 2017).
17 The acronym DEMPE (development, enhancement, mainte-nance, protection and exploitation of intangibles) was first used in2015 OECD, Aligning Transfer Pricing Outcomes with Value Cre-ation. Actions 8–10: 2015 Final Reports. OECD/G20 Base Ero-sion and Profit Shifting Project (Paris). See also 2017 OECD,Transfer Pricing Guidelines for Multinational Enterprises and TaxAdministrations (Paris, July 2017).
Tax Management International Journal4 � 2019 Tax Management Inc., a subsidiary of The Bureau of National Affairs, Inc.
ISSN 0090-4600
ties A through E perform the following functions intheir related-party transactions.
• Entity A is the customer-facing entity that housesthe IoT business leaders as well as the sales force.Entity A negotiates and signs the contract with thecustomer and markets, sells, and maintains theEdge sensor software and the equipment installedon the customer’s premises. It also provides theEdge gateway system through which the data,collected from the sensors and devices at the cus-tomer’s site, is transferred to a data storage cen-ter.
• Entity B is a related-party located in, say, China,that manufactures and sells hardware to Entity A,for use in A’s Edge gateway infrastructure.
• Entity C is the owner and operator of a data stor-age center that is located in, say, the Netherlands.It leases datacenter storage to multiple third-partycustomers. Entity C works in tandem with the IoTcustomer’s own on-premise data storage sites, ina manner that shares the data processing load andoptimizes time/cost efficiencies.
• Entity D is a data threat detection and rescuehigh-value ‘‘service shop’’ in the U.K. (for illus-tration purposes), which monitors the security andintegrity of data transmittals between entities A,C, E and the customer.
• Entity E, located in, say, Ireland, provides theCloud Platform, Cloud Applications, Data Analyt-ics software and data or information analytical ca-pabilities. The analytics that transform data andinformation into the patterns and insights that theIoT customer needs, is the ‘‘brain trust’’ of thebusiness model. The insights generated are in-stantly communicated back to the IoT customer,which applies them in real time to its productionor supply processes to improve its productivity.
In this simplified illustration of an IoT solution,there are five intercompany transactions.
• Conversion and transmittal of raw data collectedby Entity A to Entity B for storage;
• Sale of hardware, primarily sensors and fieldgateways, by Entity C to Entity A for installationat the customer site;
• Provision of security services by Entity D to En-tity C;
• Conversion and transmittal of stored data by En-tity C to Entity D; and
• Provision of security services by Entity D to En-tity E.
In addition to the five related-party transactions,there are two unrelated party transactions: one, whereEntity A collects raw data from the customer throughsensors located at the customer’s site, and the second,where Entity E transfers data insight and prescriptiveIoT solutions to the customer.
Transfer Pricing Analysis in the Direct Model
The second step of the transfer pricing analysis in-volves a characterization of the related parties after adetailed and lengthy review of its functions, assets,and risks (FAR). If the above transactions were simpleand largely unchanged during the fiscal year or overthe multiple years of the contract term, the character-ization of the related parties and subsequent economicanalyses might lend themselves to a relatively good fitwith the methods prescribed by the U.S. TreasuryRegulations §1.482 or the OECD transfer pricingframework.
Table 1 below presents the key factors consideredin the characterization process in the typical tabular‘‘FAR Matrix’’; the related parties, their key businessfunctions and headcount profiles, the risks they as-sume and assets they manage. This characterization ofthe related parties leads to the selection of the trans-fer pricing method and a profit level indicator, whichwe have also summarized in Table 1.
Table 1: Transfer Pricing Analysis in the Direct IoT Model
Related Party Entity A Entity B Entity C Entity D Entity E
Location United States People’s Republicof China The Netherlands United
Kingdom Ireland
Key Functions Sales and contractexecution.Headquarters &Business Strategy;Gatewayinfrastructuremanagement
The underlying assumption here, as is the case withall transfer pricing analyses, is that these entity char-acterizations and the underlying transactions will bestatic for the period of the analysis and all are goingconcerns. However, given that digital solutions likeIoT are inherently dynamic and non-linear and can bequickly disrupted by newer and more efficient ways ofoperating, the required assumption of indefinite stasismay not be valid.
This inherent dynamism of the IoT ecosystems isworth examining separately by introducing a PartnerModel and then using both cases to explore the distin-guishing features of the IoT ecosystem and conse-quent challenges to traditional transfer pricing meth-ods.
Transfer Pricing in the Partner ModelThe Partner ModelAs is very frequently the case, the customer may
require a more complex IoT solution to manage itstechnology, output and productivity across a growingnumber of countries, markets, and vendors. As wenoted above, one of the new ways of doing businessin a digital economy is to distribute manufacturingacross smaller, specialized units, both within the firmor between firms in a horizontal platform-based col-laboration.
In turn, the IoT company may rapidly change theconfiguration of its solution provider footprint and thesimple Direct Model may evolve into or be sup-planted by a more complicated Partner Model. In atypical industrial IoT offering, there will be multiplesuppliers and multiple entities belonging to the cus-tomer, all collaborating and providing different as-pects of the IoT solution.
Let us assume that the customer in the DirectModel is now an upstream provider of oilfield evalu-ation services to another downstream and final cus-tomer, an oil refinery. The refinery’s operations run ona just-in-time inventory basis with advanced, time-sensitive refining processes that minimize its risk andexposure to crude price volatility.
The oil refinery already works with several thirdparties in Canada and Mexico and wishes to includethese vendors in the overall solution, rather than re-placing them. Further, the new end customer has sub-stantial datacenters on premises and rather than rely-ing on and utilizing the IoT company’s off-premises
gateway, it desires to have all the data collection andanalytics be performed on premises and within veryshort timeframes. In other words, the Direct Model,with the collection of data through the gateway, stor-age and parsing at the data center, transfer to theCloud and then back to the customer creates too muchlatency for the refinery’s operating model.
Our IoT company is asked to pivot its Direct Modelto a reconfigured offering that best suits the customerwhich we refer to as the Partner Model. Both the in-tercompany and unrelated party transactions havechanged from the illustration presented in Figure 2and have been reconfigured as presented in Figure 3below.
Transfer Pricing Analysis in the Partner ModelIn the Partner Model, while there are still related-
party transactions that we viewed in the Direct Model,there are more unrelated party transactions. As op-posed to the Direct Model, the Irish entity, Entity E,no longer hosts the cloud platform or the applications.Instead, this activity is outsourced to a third-partypartner in Canada. Consequently, the ‘‘brain trust’’that was originally within the Entity E is now split be-tween itself and a third-party partner. In addition, En-tity B, the related-party manufacturer, could presum-ably be replaced by a third-party partner located inMexico. Finally, Entity A may also be required to per-form data analytics at the Edge to fulfill the customerdemand for such analytics to be performed at its ownpremises.
Tax Management International Journal
� 2019 Tax Management Inc., a subsidiary of The Bureau of National Affairs, Inc. 7ISSN 0090-4600
Table 2: Transfer Pricing Analysis in the Partner IoT Model
Related Party Entity A Entity B Entity C Entity D Entity ELocation United States The Netherlands Ireland
*Note: Bold notes are changes in the Partner model relative to the Direct FAR*
As can be seen from the FAR analysis for the Part-ner Model, the functional profile of Entities C and Ehave changed since each performs value added dataanalytics in addition to various other functions. Thisadded complexity makes it difficult, if not impossible,to apply one-sided methods such as cost plus or acomparable profits method (CPM)/transactional netmargin method (TNMM). Instead, more complexprofit split methods may be needed for a reasonableallocation of the residual profit between the variousentities contributing to value creation within the IoTecosystem—although the profit split method comeswith its own set of problems.18
While these are stylized examples, they serve to il-lustrate how the IoT solution, including its critical as-pects, can be effectively peeled off and performed by
third parties or other related parties as necessary. Theintroduction of third-party partners into the solutionchanges not only the nature of the intercompanytransactions but also the characterization of the enti-ties themselves and often, within the same fiscal yearperiod. The example also illustrates a key aspect ofIoT business models: the dynamism and ease withwhich value can be shifted between the entities in theIoT ecosystem, whether related or not, and in fact, tonew entities which were not part of the original IoTecosystem before. It also brings home the point that,in this new world of value shifts, one may have to ul-timately resort to using profit split methods from atransfer pricing perspective, regardless of the difficul-ties inherent in the practical application of such meth-ods.19
18 Eden, Lorraine, The Arm’s Length Standard: Making It Workin a 21st Century World of Multinationals and Nation States(Thomas Pogge and Krishen Mehta eds.), Global Tax Fairness(Oxford Univ. Press, 2016); Eden, Lorraine, Comments on theOECD’s BEPS Public Discussion Draft BEPS Actions 8–10, Re-vised Guidance on Profit Splits (issued July 4, 2016), CommentsReceived on Public Discussion Draft BEPS Action 8–10: RevisedGuidance on Profit Splits, Part II, at 266–269 (Paris, Sept. 8,2016).
19 Eden, Lorraine, The Arm’s Length Standard: Making It Workin a 21st Century World of Multinationals and Nation States(Thomas Pogge and Krishen Mehta eds.), Global Tax Fairness(Oxford Univ. Press, 2016); Eden, Lorraine, Comments on theOECD’s BEPS Public Discussion Draft BEPS Actions 8–10, Re-vised Guidance on Profit Splits (issued July 4, 2016), CommentsReceived on Public Discussion Draft BEPS Action 8–10: RevisedGuidance on Profit Splits, Part II, , at 266–269 (Paris, Sept. 8,
Tax Management International Journal8 � 2019 Tax Management Inc., a subsidiary of The Bureau of National Affairs, Inc.
ISSN 0090-4600
IoT CHALLENGES FOR TRANSFERPRICING
Are the current OECD Transfer Pricing Guidelinesadequate for the digital economy? How well do theexisting transfer pricing methods measure value cre-ation for the new business models of Industry 4.0?Below, we explore four distinguishing features of theIoT Business Model that challenge the relevance ofexisting transfer pricing methods for value measure-ment.
Goods, Services . . . and DataDistinguishing Features of the IoT Business
ModelIn a traditional brick and mortar value chain, the
firm’s intercompany transactions are concerned withgoods, services, intangibles, financial support, andsoftware. In the digital economy, data and informationmay often be the only products being exchanged be-tween related parties. As shown in the IoT case stud-ies, data are transformed into millions of formless(0,1) packets that are constantly digitally transferredbetween the firm’s legal entities, its vendors, co-developers, and customers. The data or informationcan either be unique and proprietary (collected in ahighly customized IoT solution) or simply obtainedfrom the public domain. An additional complicatingfactor is that non-proprietary data can be uniquelytransformed during the intercompany exchange to ac-quire a unique profile and a market price. Data is anon-tangible but may not be an intangible asset in ev-ery cross-border exchange.
Transfer Pricing ChallengesWhen the transaction involves the exchange of eas-
ily identifiable, separable goods and services acrossthe value chain, the transfer pricing practitioner canlook for identical or comparable products (goods, ser-vices, intangibles) that are traded in markets whereprices and terms may be publicly reported. The separ-ateness and identifiability of products and servicesmakes possible, a matching of characteristics andidentification of comparable uncontrolled prices(CUP) or comparable uncontrolled transactions(CUT). Data and information on the other hand, donot have market prices per se, making the CUP orCUT, the most reliable of the transfer pricing meth-ods, impossible to apply.
Circularity and Value ShiftsDistinguishing Features of the IoT Business
ModelOne of the unique aspects of IoT value chains is
that data, even if it were a commodity at the onset,
becomes increasingly valuable as it accumulates. Thisis because a larger number of patterns and disso-nances can be detected in larger volumes of data andthat information can be used to derive value-added so-lutions for the customer. The larger the volume of‘‘big data,’’ the greater the scope of marketable in-sights that can be generated. Further, the transferenceof data cannot be captured as a flow along a single lin-ear value chain from the producer to the customer. Itis circular—data is first collected from the customerthrough sensors and edge devices and is then trans-ferred through the device and software in the distrib-uted core, to the cloud, and then back again to the cus-tomer as insights and solutions to the customer, whouses it as inputs and sends output data back to theEdge, Core, Cloud, etc. Where the customer and theMNE have operations around the world, this circularexchange of data can occur 24/7 in a manner some-what similar to the 24-hour global trading models forfinancial institutions20
This circularity of the IoT data/insight exchangemeans that the ‘‘final’’ end-product is not the outputof the transactions between entities in either the Di-rect or Partner Model. With every iteration of transac-tions through the value chain, the accumulation ofdata and speed of insight generation grows, generat-ing more value with each iteration. As larger volumesare processed faster, deeper and more marketable in-sights are also generated as time goes on.
As data exchanges increase and as data transformsinto information and insight, as the related-party andthird-party entities perform multiple and intercon-nected functions in relation to the flow and processingof data, how does the practitioner and the tax author-ity, as a precursor to evaluating nexus and applyingprofit allocation, pin down which entity is transferringroutine data and which entity is transforming the rou-tine data into valuable insight? Clearly, the digitaleconomy throws off challenges even at the outset ofthe standard transfer pricing analysis—identifying the‘‘simpler of the tested parties’’ to apply one-sidedtransfer pricing methods.
Circularity also creates an interesting dilemma interms of when the ‘‘sale’’ takes place. While account-ing standards determine revenue recognized over thecontract period per specified milestones, does circu-larity, from an economic perspective, imply that valueis continuously generated for the customer? If this isthe case, should we rethink how to structure thesecontracts and how customers pay for them?
2016).
20 Eden, Lorraine, Taxing Multinationals: Transfer Pricing andCorporate Income Taxation in North America, at 574–578 (Univ.of Toronto Press, 1998).
Tax Management International Journal
� 2019 Tax Management Inc., a subsidiary of The Bureau of National Affairs, Inc. 9ISSN 0090-4600
Transfer Pricing ChallengesThere is no disagreement that digital business mod-
els in the 21st century have a much higher degree offunctional complexity than hardware manufacturingor services provision in the 1980s and 1990s. Thatsaid, there is disagreement about whether the ‘‘serviceprovider’’ or ‘‘intangible property owner/developer’’characterizations still apply when the intercompanyproduct being transferred is data or information, orwhether the cloud-based applications developer isalso a manufacturer or a principal innovator of ana-lytical insight. A related question is whether the‘‘value’’ is created and controlled by the headquartersor the entities in the transaction chain that are in-volved in the data flows and transformation. Unliketraditional business models, in an IoT solution wherevalue creation is not a linear process, it is not clear ifthere is any single principal entity or brain trust whichcontrols the creation of value. Even if there were twoor more principal entities, the same dilemma exists.
Indeed, at what point in the IoT value chain doesthe data acquire value? How could that value be mea-sured? Is it a routine or non-routine intangible? Inmost of the IoT solutions, data is transferred betweenlegal entities at lightning speed. If the value is createdwhen the data is transformed into a ‘‘prescriptive so-lution’’ for the customer, the entity taking on the ac-tivities and risks of this transformation should be theprincipal value-driver or the entrepreneur. To set theprice of the data transferred to this principal entity, itis necessary to characterize the related-party trader asa routine manufacturer or service provider, which itmay not be. It is hard to pin down the exact momentwhen the data acquires value, which in turn makes thelegal entity that transfers or processes that data harderto characterize from a transfer pricing perspective.
Speed of Technological Change and FunctionalityDistinguishing Features of the IoT Business
ModelIoT is fundamentally designed to be smarter and
more flexible over time. The digital economy is char-acterized by high technological obsolescence wherefirms compete fiercely to be first to market with alower-cost solution that delivers higher customer sat-isfaction. This in turn means that functions, risks andthe intangible property developed by entities canchange as the several cycles of data collection-transformation-feedback occur and more efficientways of delivering value to the customer emerge.
An example of this in the IoT space is the recentemergence of distributed analytics. Distributed analyt-ics is the technology and capability to shift the ma-chine learning and data analytics from the cloud to theedge so that sensors and devices that were previouslydeployed only to collect data are now also performing
certain analytical tasks ‘‘in the moment.’’ The outputthat previously came from the Cloud entities are nowbeing generated by entities that own devices on theEdge (viz., sensors, mobile phones, laptops, etc.). Dis-tributed analytics is attractive for certain customerswho need to minimize latency and increase responserates for time-critical IoT applications such an autono-mous vehicles or robot-supplemented surgeries. In theDirect Model, this would mean that the insight gen-eration functions can be shifted from Entity E to En-tity A or from Entity C to A. Moreover, the shift mayoccur within a single fiscal year depending on howquickly the customer’s requirements change. As func-tions shift across entities, so will the risks of manag-ing the costs of technological or market failures.
Transfer Pricing ChallengesIf the related-party’s functions change from routine
data collection/processing service to analytics and in-sight generation in a short period of time, even withinthe same fiscal year, what is a reliable characterizationof the entity? A dynamic functions-assets-risks (FAR)profile presents the taxpayer with the very simplepractical challenge of preparing annual transfer pric-ing documentation that describes the entity’s func-tions for the first say, six months as significantly dif-ferent from the next six months of the year. When allintercompany transactions must be priced at arm’s-length, the functional shifts can mean switching froma cost plus method to a TNMM/CPM or even consid-eration of a profit split.
Functional shifts challenge the very idea of valueattribution to specific legal entities based on theirfunctional profile, a fundamental aspect of transferpricing analyses. The key question for taxpayers is notwhether to recharacterize entities constantly and re-price the flow of goods, services and data but whetherthese dynamic business models will be well under-stood and accepted by tax authorities.
If the key functions, assets, and risks change everytwo or three years, or even every year, taxpayersshould weigh the costs of impending controversyagainst a more practical approach of pegging a targetoperating margin based on comparables that ‘‘do itall’’—from sales and marketing to data analytics toasset management and sale—rather than trying to as-sociate the entity’s returns with too narrow a set ofmore functionally comparable companies, that withfunctional shifts, may become redundant and stir upmore rather than less controversy.
To be fair, the emergence of Industry 4.0 does notmean that every digital economy firm needs a brand-new set of transfer pricing rules. For the majority ofthe Going Digitals, the traditional cost plus method orTNMM/CPM can be applied to separable routinefunctions such as data transference services or hard-ware production and supply. These transactions within
Tax Management International Journal10 � 2019 Tax Management Inc., a subsidiary of The Bureau of National Affairs, Inc.
ISSN 0090-4600
the digital value chain are fundamentally like those ina brick-and-mortar environment. They are at the lowend of functional complexity even if there is no‘‘more complex’’ counterparty that controls or man-ages their activities and risk taking.
Conversely, traditional transfer pricing methods be-come less and less useful as complexity increases andrelated parties cannot be easily characterized, eitherbecause their functions and risks are evolving andchanging as the IoT solution matures, or because thevalue is in what they transact (big data) and not intheir activities.
The profit split method, while more cognizant ofdistributing value among complex entities, is stilltethered to some level of comparable benchmarkingfor routine functions. In addition, the actual imple-mentation of a residual profit split method involves avariety of decisions and complications that may ren-der it worse than a TNMM/CPM analysis added to anintangible asset valuation.21
Control, Decentralization and CooperationAmong Related Parties
Distinguishing Features of the IoT BusinessModel
An IoT offering is more akin to a loosely connectedchain of value shops than a traditional value chainwith linearly dependent set of related parties perform-ing sequentially interconnected functions. The notionof which party controls another and with what eco-nomic substance is less amenable to a DEMPE-typecharacterization. IoT providers will typically havemany groups of highly skilled data scientists and soft-ware engineers distributed across many different ge-ographies (and legal entities) that simultaneouslywork at transformation and analyzing big data. In ad-dition, the interaction between these groups and valuecreation is often too fluid to fit them within a tradi-tional ‘‘principal-agent’’ framework.
Transfer Pricing ChallengesIn fact, in most IoT solutions, there is no traditional
‘‘principal-agent’’ relationship or contractor-subcontractor arrangement between the parties,whether related or unrelated. Entity C is not providingdata security on behalf of Entity E, nor is the redistri-bution of analytical capabilities from E to Entity A onthe Edge some type of sale or license of Entity E’s in-
tangible property to A. Without a large capital invest-ment and only a relatively small but scalable invest-ment in human capital, functions based purely onworkforce competency can quickly migrate from onejurisdiction to another.
For example, in the Partner Model typical of IoTofferings, without a traditional control and functionaldependency and without a ‘‘simpler’’ party, it is diffi-cult for the transfer pricing practitioner to embrace theapplication of TNMM/CPM. While the number of dis-creet intercompany transactions has certainly dimin-ished, more third-party vendors and value shops arenecessary to work a complex solution. Even if thereare no intercompany transactions, related parties maystill influence their siblings’ financials or operationswithin the overall value chain. It could be argued thatthe interactions with third parties provide the besttransactional comparables for gross profit or operatingprofit margin setting. This is true only if the uncon-trolled parties and related parties transact on a rela-tively frequent and stable basis. When multiple cen-ters of excellence for data analytics or cloud hostingexist across related and unrelated parties, using a lim-ited duration gross margin comparison could be lessaccurate than an imperfect application of a CPM/TNMM.
CONCLUSIONAre the current OECD Transfer Pricing Guidelines,
including the recently issued programme of work in2019 OECD,22 adequate for the digital economy taxa-tion? How well do the existing transfer pricing meth-ods measure value creation for the new business mod-els of Industry 4.0? In the first article in this two-partseries on this topic, we explored the old and new firmsin the digital economy and reviewed the OECD’sBEPS project focusing on BEPS Action Items 1 and8–10.
In this second article we provided an overview ofIoT and outlined the Direct and Partner BusinessModels. We evaluated each model using a traditionaltransfer pricing analysis, and explored four challengesthat these new business models create for transferpricing: data-based related-party transactions; circu-larity of and value shifts in the IoT data/insight ex-changes; the speed of technological change and func-tionality; and difficulty in characterizing control, de-centralization and cooperation among the relatedparties.
Our IoT case study has attempted to illustrate someof the key challenges of applying existing transfer
21 Eden, Lorraine, The Arm’s Length Standard: Making It Workin a 21st Century World of Multinationals and Nation States(Thomas Pogge and Krishen Mehta eds.), Global Tax Fairness(Oxford Univ. Press, 2016); Eden, Lorraine, Comments on theOECD’s BEPS Public Discussion Draft BEPS Actions 8–10, Re-vised Guidance on Profit Splits (issued July 4, 2016), CommentsReceived on Public Discussion Draft BEPS Action 8–10: RevisedGuidance on Profit Splits, Part II, at 266–269 (Paris, Sept. 8,2016).
22 2019 OECD, Programme of Work to Develop a ConsensusSolution to the Tax Challenges Arising from the Digitalisation oftheEconomy, OECD/G20 Inclusive Framework on BEPS (Paris).
Tax Management International Journal
� 2019 Tax Management Inc., a subsidiary of The Bureau of National Affairs, Inc. 11ISSN 0090-4600
pricing frameworks to an IoT ecosystem. However,IoT is just one among many digital business modelsthat are emerging as we move into Industry 4.0.Transfer pricing analyses of other digital businessmodels are also needed before we can start to developa robust understanding of our current transfer pricingmodels and methods in Industry 4.0.
We conclude that more work is needed to ‘‘lift theveil’’ on transfer pricing in the digital economy. A bet-
ter understanding of how the current transfer pricingrules apply to digital economy models is needed be-fore governments should make additional changes tointernational tax and transfer pricing policies. As wemove into Industry 4.0, we hope that our analysis willcontribute to the ongoing debate about reforming therules with a renewed focus on updating the arm’s-length standard for measuring value and allocatingprofits in the digital economy.
Tax Management International Journal12 � 2019 Tax Management Inc., a subsidiary of The Bureau of National Affairs, Inc.
