Aligning China R&D Arrangements and Transfer Pricing in a Post BEPS World

James Yi Min Zhao and Victor Zhang | Deloitte

Economist James Zhao, Ph.D., tax partner of Deloitte, and Victor Zhang, tax director of Deloitte, have published the captioned piece in The Bureau of National Affairs on May 24, 2018.

What is the appropriate structure in China for a multinational group’s research and development (RD) activities, and intellectual property (IP) ownership, to accommodate business dynamics, entitlement to tax incentives, whilst fitting into the group’s global transfer pricing arrangement in the post BEPS era?

This question is becoming more and more important for multinational companies operating in China.

R&D activity and the ownership of IP is often a core issue in multinational tax and finance considerations. In a post base erosion and profit shifting era, R&D arrangements and relevant IP ownership structure is expected to be under closer transfer pricing scrutiny by tax authorities. One significant focus is the contribution of development, enhancement, maintenance, protection, and exploitation (DEMPE) functions of the jurisdictions in the value chain.

On the other hand, many jurisdictions offer R&D incentives to multinational companies to encourage technological innovation.

There has been a gradual change in R&D activities in China, partly due to the growth of local manufacturing capability with increased accommodation to local and global needs from the demand side. This is also partly due to the Chinese government's encouragement of local R&D activities and innovation.

This change in R&D dynamics in China further complicates the set-up of transfer pricing arrangements under the post BEPS era.

This paper presents an overview of the China’s R&D tax incentives, transfer pricing regime, and recent R&D developments in China. The paper then presents options available to MNCs to structure their local R&D activities and IP ownership.

This article aims to present a broad framework for MNCs to analyze the global, versus China R&D arrangement and IP ownership structure.


R&D Tax Incentive and Transfer Pricing Regime

With the growing emphasis and encouragement of the Chinese government on innovation and the 2025 national strategy of "intellectual manufacturing," the Chinese tax regulations offer various tax incentives to encourage local R&D activities, including:

  • High and New Technology Enterprise (HNTE). The local Chinese entity must retain ownership of the technology. There is a local enterprise income tax rate of 15 percent (versus the 25 percent statutory enterprise income tax rate);
  • Technology Advanced Service Enterprise (TASE). The local entity's service must fall within the recognized scope, and 35% of income must be from offshore outsourcing business. The local enterprise income tax rate of 15 percent applies; 
  • R&D Expense Super-deduction. For qualified R&D expense, a 50% super-deduction is allowable in addition to the actual expense deduction. If the expenditures are capitalized as intangible assets, cost bases of the intangible assets will equal 150% of actual costs that are allowable for amortization purposes; and
  • Industry-specific incentive applicable to the software and integrated circuit business.

China's transfer pricing regulations were updated in 2016 and 2017, incorporating many of the OECD's BEPS Action 13 and Action 8-10 recommendations. The State Administration of Taxation (SAT) released new rules on Contemporaneous Documentation (Bulletin 42) and Advanced Pricing Arrangements (Bulletin 64) in 2016, and new rules on Special Tax Audit Adjustment and Mutual Agreement Procedure (Bulletin 6) in 2017.

Specifically for intangible property related considerations, China’s tax authority echoes the OECD's view that intangible returns should be allocated to the entity performing DEMPE functions, but the tax authority also gives additional emphasis to local promotion as being an important IP function.

In the U.N. Transfer Pricing Manual, the SAT re-emphasizes its attention to the matters of location specific advantages and the local entity's contribution to the MNC group's intangible property. SAT uses an example that the local entity may not constantly pay a technology royalty to its overseas parent company for a long term (as the licensed technology value may be reduced over time), and if the local entity has developed intangibles shared with other group companies, the local entity should be entitled to a return on its intangibles. See Paragraph D., China Country Practice chapter of the U.N. Transfer Pricing Manual (2017).


R&D Activities in China

Previously, it was more common for MNCs to characterize their China R&D operation as contract R&D and all of the legal and economic ownership of any IP would be held by an overseas entity. This is partly because the local R&D activities are less significant and mostly under close direction and supervision from overseas. Depending on the specific industries and companies, our recent experience shows that now MNCs' R&D operations in China have become more diversified, and in some cases, it will be probably viewed as more than a pure contract R&D set-up.

In recent years, some MNCs' R&D global activities have been performed in a relatively decentralized manner for accommodation to local market needs. For example, R&D teams from different regions, often with China as one of the hubs, now focus on different priorities with sufficient local discretion in initiating projects and R&D decisions, instead of a local R&D team following instructions mostly from its headquarters. The team in China will often exchange ideas and solutions with overseas R&D hubs or their headquarters for mutual benefit. Therefore, it is difficult to strictly require all of the DEMPE functions to be centralized in one legal entity, for example, the local China R&D teams play a significant role generating outcomes benefiting both the China and overseas operations.

While many MNCs traditionally prefer to have a centralized IP structure because of a clear-cut transfer pricing system, following the "business first" principle, it is unlikely that the tax and finance function would ask the R&D organization to make a substantial change in their work procedures. It therefore requires MNCs to explore some alternative R&D arrangement options by balancing the group's IP strategy and transfer pricing policy, against the local R&D incentives and actual business needs. In the rest of the paper, we will examine feasible options.


Potential R&D and IP Ownership Structure

Contract R&D and Technology License Option

In a common arrangement, MNCs structure their China R&D as contract R&D services for the overseas IP owner and often compensate the local R&D activity via a cost plus service charge. The overseas IP owner then licenses the IP to the operational entities in China and globally, and collect a technology royalty.

While this is still a relatively common structure and it follows the centralized IP strategy, some MNCs are concerned about whether this option is sustainable in the long term, particularly if there are significant functions and DEMPE substance in the local entity.

On the other hand, China SAT is very keen to understand the MNC group's overall value chain and R&D arrangement when making transfer pricing assessments. There is a potential risk of the contract R&D service relationship being re-characterized under the view that the China entity is the economic owner of the locally developed IP by examining the local DEMPE functions.

Therefore the contract R&D service option could be suitable for those with robust DEMPE functions in the overseas entities, and the overseas entities could excise control over the key R&D risk related to the China R&D projects. When setting up the contract R&D arrangement, MNCs would also note that the location specific advantage is considered in the comparable analysis. Contract R&D is an area of focus for the SAT to apply the location saving argument.

However, under the contract R&D service relationship, because the China entity does not have any IP ownership, it could not qualify or China HNTE status, although under certain special arrangements where the China entity is a pure service entity, it could assess whether it is possible to apply for the TASE status.

In some cases, when the local entity does have significant DEMPE functions and contributes to the ownership of IP, MNCs may value the locally developed IP and then sell back to the overseas owner for the purpose of achieving a globally centralized IP ownership. This is referred to as the "Valuation and Sell" option. Again, this alternative renders the local entity being unable to enjoy local R&D incentives such as the HNTE status.

Local Ownership Option (with Possible Cross-licensing)

Many Chinese subsidiaries of MNCs performing local R&D activities consider qualifying for HNTE status to enjoy the tax incentives. If the MNC group still intends to centralize the core IP out of China, the China HNTE status requirement will lead to a potential conflict. The prevailing China HNTE rules require that the local entity has the IP that "provides core support to its product/service." See GuoKeFaHuo [2016] No. 32.

However, often from a transfer pricing perspective, a MNC’s China entity does not own the group’s core IP, and the Chinese entity may still may be positioned as a routine company with routine return.

In order to reconcile the core IP ownership required by the HNTE rules and the routine return transfer pricing policy, MNCs sometimes take the approach of categorizing the group IPs into different levels:

  • Tier 1 IP: the foundation of technology, normally centralized by the group-level IP owner; and 
  • Tier 2 IP: the application IPs which are developed based on the Tier 1 IP, and could be owned by local countries.

In this situation, the China entity could own the Tier 2 IP in the HNTE application. Given the ambiguity of the HNTE qualification standards and the judgment call by the government agencies involved in the review process, it is possible for the China entity to obtain HNTE status even it does not own the Tier 1 IP of the group. Meanwhile, the China entity's routine return is supported with the argument that the economic contribution of Tier 2 IP is substantially less than Tier 1 IP. In such a case, the China entity would need to make an overseas royalty payment for the use of Tier 1 IP.

This strategy seems to provide an ideal answer to MNCs in terms of enjoying the local HNTE incentive and consistently following the group's IP strategy and TP policy, if the facts and circumstances support the strategy.

However, with the HNTE status, the local tax authorities often pay more attention to the transfer pricing policy. The tax authorities will probably require a higher than routine return in an actual transfer pricing audit. A robust documentation effort is required to convince the specialized SAT anti-tax avoidance team that the local IP ownership is much less valuable than the global core IP ownership and local Tier 2 IP ownership does not entitle to a higher than routine return. See "Transfer pricing considerations for companies seeking 'High and New Technology Enterprises' Status in China," Vol. 23 Transfer Pricing Report 313, June 12, 2014.

In addition, MNCs would also need to pay attention to innovation results potentially creating an impact beyond the particular jurisdiction’s borders because of globalization. For example, an R&D project conducted in country A could possibly be used in country B's business, and vice versa. That is to say, the China entity might own certain IP that is used by affiliates in other countries, with the China entity also making use of the base IP that is owned by the overseas principal. In this case, it may require a complex cross-licensing agreement among the group entities with implications for withholding tax and turnover tax. These complications mean some MNCs will be less willing to take the risk of claiming HNTE incentives even though they have local R&D activities, because it could substantiate the local IP ownership and complicate the global transfer pricing arrangement of the MNCs.

Cost Sharing Arrangement (CSA) Option

China’s SAT introduced the CSA rules in Bulletin 2 in 2009, which have not been substantially updated since the OECD BEPS Actions. However, the SAT presents its view of IP and economic analysis in other newly released transfer pricing rules that could provide some guidance regarding China cost sharing arrangements. OECD, U.S., and China positions on cost sharing arrangements are:

CSA Rules

OECD (2017)

U.S. (2011)

China (2009, partial update in recent years)

Applicable scope

Development, production or the obtaining of IPs, tangible assets and service (though the guidance is mainly about intangibles)

IP developments

IP developments and intra-group service of group procurement and group marketing

IP return allocation to participants

Perform DEMPE activities and manage and control risks

No specific guidance on IP activities

Expect similar to OECD

Assessment of current contribution

Be assessed based on  value at the time of contribution; only in limited cases cost could

be used as a practical means

Be assessed based on  cost

Expect similar to OECD (SAT would likely require the location specific advantages considered in the value contribution)


Under a CSA, the China entity would co-own the relevant locally developed IPs with its overseas affiliates, while the base IPs would likely be with the group IP owner. The CSA participants will share the IP development contributions based on the respective expected benefits, i.e., the China entity is entitled to the sole IP usage and benefit in the China business operation, while the IP usage and benefit in the overseas operation will belong to the overseas affiliate(s).

This option could provide MNCs with some flexibility for the R&D arrangement because the base IP is centralized with the group IP owner, whilst the China entities could perform local R&D activities based on the global base IP and with on-going contributions from overseas. Also, the China entity and the overseas entities could make periodic adjustments (e.g., on an annual basis) of contribution payments based on any update of the expected economic benefits, which could avoid the alternative complex cross-licensing arrangement. The China entity could, with the IP ownership under the CSA, also apply for the HNTE status and benefit from the local R&D incentive.

However, unlike the U.S. and some of the other countries, China still does not have detailed CSA rules. It is still a relatively advanced arrangement for the China tax authorities where certain practical aspects including turn over tax and WHT preferential treatments need to be further clarified.

Even if the taxpayer does not need to obtain an advanced approval for concluding a CSA, it needs to submit a copy of the CSA to the tax authority for disclosure purposes. In addition, the local entity with the CSA would need to prepare and submit upon request the contemporaneous CSA special file documentation on an annual basis.

To increase certainty, MNCs may also consider applying for an advanced pricing agreement to increase tax certainty, although management would have to consider many other factors before making an APA application.

Overall, this option may be able to accommodate the requirements for many MNCs' global R&D arrangement and IP ownership structures. The transfer pricing arrangement under this CSA option would also require robust economic analysis because it is likely to be under the spotlight of tax authorities, but it could well be advantageous given that the China entity does co-own the locally developed IP and hence maximizes the likelihood of enjoying relevant R&D incentives such as HNTE status.


Concluding Remarks

In the post-BEPS era, MNCs face increased scrutiny of their global IP structure and R&D activities.

The Chinese tax authorities have been placing great emphasis on transfer pricing involving IP with recent Bulletin No. 42, No. 64, and No. 6 where the BEPS recommendations are largely incorporated.

The R&D dynamics have been changing in China with the government's strong encouragement of innovation under various incentive programs and strategic programs such as the 2025 national strategy of "intellectual manufacturing."

We have observed more-and-more local R&D activities in China by the MNCs, depending on specific industries and specific business cases. This has created challenges for MNCs on how to arrange their IP ownership structures. In particular, the common centralized IP ownership structure using contract R&D service arrangements could face potential challenges whilst foregoing the R&D incentives. On the one hand, tax authorities cite the BEPS recommendations and new Chinese transfer pricing rules to focus on assessing the local contribution to DEMPE functions and require a share of the return attributable to the IP. On the other hand, there have been increased local R&D activities in China for business purposes.

We have analyzed the various options available to MNCs for structuring their local R&D activities and IP ownership in China. There is no perfect option available to fit all MNCs given their diversified business cases and objectives. CSA appears to be a good option for management to address the above challenges whilst enjoying R&D incentives under certain scenarios. Although the rules on CSAs in China are not as clear as business professionals would like, we expect the Chinese tax authorities to be potentially more open to the CSA arrangement, and continue to relax relevant administrative requirement for CSAs in the coming future.

Economist James Yi Min Zhao, Ph.D., is a transfer pricing tax partner, and Victor Zhang, is a transfer pricing director, with Deloitte Shanghai, China. The authors’ views are their own, are not Deloitte's official position, nor do they constitute advice to any companies.

Did you find this useful?