Innovation-based tax incentives offer ample tax planning opportunities. Still, patent law limits applicability to digital business models. Thoughtful IP allocation across jurisdictions is key.

With the introduction of innovation-based tax incentives in 2020, Switzerland made an important step to remain extremely attractive for innovative companies. While a legal entitlement to R&D tax incentives provides planning opportunities for taxpayers, fulfilment of certain criteria may depend on the specific innovation and business model.

The new tax incentives are based on two pillars:

  1. Innovation that ultimately results in patents can unlock benefits from the patent box. Net profits from patents and products with embedded patents may be partially tax privileged, while a connected potential entry tax should be thoroughly managed and mitigated.
  2. For qualifying research and development costs from ongoing projects, companies may, upon request, benefit from an R&D super-deduction over and above the incurred expenses that are commercially justified and recorded in the accounts (super-deduction for research and development costs).

Limitations to the patentability of digital innovations

The rapidly changing economy challenges existing legal frameworks on a global scale. Policymakers and experts around the world are faced with the question whether existing laws are ready to respond to the promises and risks of new technologies. There is an inherent difficulty for legal intellectual property (IP) and R&D frameworks to grasp the digital technologies, as they are naturally limited by the underlying IP laws currently applicable. Legal uncertainties in connection with the patentability of digital innovation result in uncertainties for the taxpayer. In order to determine potential benefits of a tax planning strategy involving the patent box in the digitalized economy, it is therefore key to first understand the highly disputed questions and hurdles of the underlying patent law and practice concerning digital innovation.

In European jurisdictions, digital innovation currently is only patentable in the form of computer-implemented inventions (CII). CII are inventions of which one or more features are realized (“implemented”) by software on a computer. This amalgamation of digital and physical matter is however only patentable, when the soft- and hardware represent a technical solution to a technical problem. As a result, digital innovation needs to overcome two hurdles: The presence of a computer or computer-network and a successful technicality analysis.

Business models in the digital economy however shift from selling physical products to selling digital products and services, which do not qualify as CII and are therefore not patentable, as seen with many fintech-innovations. Thus, digital innovators will often not be able to reap the benefits offered by the patent box in a narrow sense, i.e. a box which excludes software as qualifying income, as it is the case in Switzerland.

Deduction of qualifying R&D activities resulting in digital innovations

The Swiss R&D super-deduction allows for increased tax deductions by multiplying qualifying actual R&D expenditures incurred by up to 150% (at cantonal discretion, detailed regulations apply). In contrast to the reward-based approach of the patent box, the R&D super-deduction follows an input-based approach. More specifically, R&D carried out by the taxpayer does not need to result in a patent in order to qualify for the super-deduction and thus opens a much broader ambit of applicability.

This tax incentive is therefore not restricted by underlying IP laws and not impacted by the hurdles and challenges of patenting the results of digital innovation. From the taxpayers’ perspective this means that even if they cannot profit from the privileged taxation within the patent box, the expenditures incurred by their digital innovation efforts may qualify for the super-deduction and the innovator can profit from this instrument without being dependent on the patentability of the final product. Deductibility is tied to the scientific definition of research and development, which is typically interpreted based on the Frascati-criteria[1], meaning that any R&D activities must be, cumulatively:

  • novel
  • creative
  • uncertain
  • systematic
  • transferable and/or reproducible
ey-switzerland-blog-Tax incentives

The patent box and the R&D super-deduction are introduced and interpreted largely inconsistently across Switzerland. Each measure’s different timing of applicability is another crucial point to consider when setting-up the tax strategy, as the same do not align within the digital product life cycle (i.e. development phase not matching with the commercialization) and there is no benefit-carry-forward mechanism in place yet. The well-established Swiss advance tax ruling practice proves helpful in this regard, as it allows to agree with one or more canton(s) on tax parameters of projects before implementation in order to achieve legal certainty.

Deductibility of artificial intelligence (AI)-driven innovation

The R&D super-deduction may even offer room for tax planning in cases where it is no longer a human inventor driving innovations but rather a computer, equipped with AI. AI is often used to describe machines (or computers) that emulate “cognitive” functions associated with the human mind, such as “learning” or “problem-solving”. AI-driven innovation mainly relies on genetically programmed computers, which automate technical high level problem-solving by improving upon a set of candidate solutions and parameters of known performance.

The super-deduction relies on personnel expenditures as a calculation basis and the Frascati-criteria define research and development as a process realized by a human inventor which may not yet be considered old-fashioned but may soon be anachronistic. Expenditures directly incurred by or for the AI in use, i.e. energy or material cost, are therefore not to be considered as a basis for the super-deduction as such but are globally recognized by a fixed modest uplift of max. 35% granted on top of the personnel expenditures. As things stand today, AI driven processes still necessarily induce human expenditures, namely for the development of the AI itself as well as setting the parameters and selecting the candidate set of solutions to be utilized by the AI in the innovative process. At least these human expenditures may bear potential for tax planning in cases of an AI-inventor.

Worldwide R&D incentives

Tax competition — driven by governments vying to attract and sustain economic activity in their jurisdiction — is expected to generate further global tax changes in 2021. As countries formulate plans to reinvigorate their economies in the wake of COVID-19, we expect many to implement new, expanded or specially focused R&D incentive programs.

In this fast-changing R&D and innovation incentives landscape, the EY Worldwide R&D Incentives Reference Guide has been an invaluable tool for taxpayers who need to stay abreast of changes in this space, especially if they are considering new or expanded investments in R&D. It includes a description of available benefits as well as an overview of the incentive application process and it covers in detail the eligibility and IP jurisdictional requirements. Where available, patent and innovation box regimes are also highlighted. You can access the complete guide here.

The OECD member states harmonized their innovation-based tax incentives to an extent that they agreed on implementing minimum substance requirements in their respective IP box regimes. Contrary to Switzerland, various jurisdictions allow income from software to qualify for an IP box, e.g. by integrating copyrights as qualifying IP. The Netherlands, for example, grant access to their “innovation box” for any IP that can be classified as software.

ey-switzerland-blog-Tax incentives

Recommendations for Swiss based companies

Multinationals are well advised to define an overall strategy on where to best allocate their IP and – if feasible from an operating perspective – might even consider assigning them on multiple jurisdictions to exploit further advanced resp. more favorable foreign tax regimes as a matter of reducing the overall group effective tax rate. Furthermore, cross-country and foreign R&D incentives should be evaluated and considered in the design of the operating model and execution of R&D. This may prove particularly important for start-ups in the digital space, which typically already operate in multiple markets at early development stages.

To evaluate and eventually implement opportunities based on innovation-based tax incentives in Switzerland recommended steps include, inter alia:

  • Qualification of activities and products that might be entitled for patent box and/or R&D super-deduction
  • High level assessment on cost-benefit relations
  • In-depth analysis of key projects and products
  • Development and setting of a holistic tax strategy
  • Evaluation and obtainment of clearance in all jurisdictions involved, e.g. in Switzerland via the instrument of advance tax rulings

EY’s subject matter experts are ready to support you. Please contact us anytime.

[1] Established by the OECD in the “Frascati Manual 2015: Guidelines for Collecting and Reporting Data on Research and Experimental Development”