Industry 4.0 or the internet of things may already sound like yesterday’s concepts in the era of ever-accelerating technological advancements. Today, the conversation is dominated by AI-centric and AI-adjacent expressions. Yet, regardless of the buzzword of the moment, one fundamental enabler of innovation remains unchanged: data.
Multinational groups have been investing for decades to collect, structure, store, exploit, and protect data with the aim of creating long-term business value. However, there is no international consensus on how transfer pricing should address the value created through effective data management.
Particularly, in the industrial space, there is neither comprehensive guidance nor extensive case law and doctrine regarding this topic. Yet the use cases for industrial data within a related-party environment are extensive, ranging from predictive maintenance and quality control to supply chain optimisation and energy management, among many others.
With companies making smart factories a reality and harvesting data at scale to improve performance, it is important to define an analytical framework that is both practical and technically sound.
Smart factories and the transfer pricing challenge
Becoming a smart manufacturer requires a number of steps:
Defining a business case and strategy;
Digitalising processes and equipment through sensors, computers, and servers;
Deploying the right cybersecurity and system integration capabilities;
Designing advanced analytics; and
Identifying optimisation measures that can be implemented automatically or subject to human assessment.
There are significant investments to be made before benefits can be reaped. These may arise both at the factory level – such as sensors, actuators, and cybersecurity – and at the central level, including system integration, design, data consolidation, and analysis. Change management of people involved in the project should also be addressed at both levels.
Centralised versus decentralised manufacturing structures
In a centralised structure – for example, a contract or toll manufacturing around a principal structure – the transfer pricing aspects of industrial data are relatively straightforward. Local investments are initiated at the request of the central organisation and are indirectly financed by it. The funding of these investments occurs over time through the contract manufacturing price or tolling fee, reflecting the depreciation and amortisation of the underlying assets. The central organisation retains overall control of the project, assumes the investment risk, and, ultimately, derives the benefits – or losses – of that smart manufacturing strategy.
However, the analysis becomes more complex when manufacturing entities operate with a higher degree of autonomy.
In a decentralised structure, where multiple manufacturing subsidiaries function as fully fledged operations, the transfer pricing impact of industrial data is more difficult to assess, particularly when a central headquarters or digital hub initiates and funds the investments. The transformation into a smart factory affects both revenue (top line) and profitability (bottom line). These impacts can arise in the short term or develop over a longer horizon, and not all use cases may be clearly defined at the time the initial local investments are approved and implemented.
In practical terms, the transformation of a fully fledged plant into a smart factory introduces a new element into the core manufacturing value chain. This new component is deployed and operated through a combination of local and central efforts and investments.
Choosing the right transfer pricing model
Depending on the functional and risk profile of the fully fledged parties and how the transformation is organised, there may be at least four different transfer pricing models that could be implemented:
Profit split – where the centre (headquarters or digital hub) and the fully fledged manufacturers jointly decide to embark on the smart factory journey, their respective unique contributions to the value chain may lead to a profit split approach, often implemented as a residual profit split. While this may be technically appropriate, it also raises several complexities, including how to isolate the benefit from the smart factory itself, how to determine the appropriate profit or loss split, and how to identify the entities that should participate in that split. In practice, the agreed allocation may be reflected economically through mechanisms such as a royalty or licence-type fee paid to the centre. Given that manufacturing entities are often located in jurisdictions with complex transfer pricing environments, applying the profit split method may prove to be a source of regular controversy.
Cost contribution arrangement (CCA) – if the group’s primary objective is to develop an intangible asset related to operating smart factories using high‑quality industrial data, a CCA may be an appropriate option. A CCA can help preserve the existing functional profile of the participating entities and may offer a comparatively more straightforward framework for implementation and ongoing administration than a profit split model. That said, a CCA is not without challenges. Practical issues include managing the entry and exit of participants, proving that the benefit from the CCA is proportional to the participation, and identifying qualifying costs. Nevertheless, if designed and executed properly, a CCA can be structured to support an arm’s-length allocation of costs and benefits among the parties.
Return on assets (RoA) – if the priority is to channel the benefits of industrial data to the centre while preserving the fully fledged status of the manufacturing plants, the group may consider compensating the plants with a routine RoA for the local investments required to support the smart factory transformation. The distinction between ‘traditional’ and smart factory investments in order to remunerate the first with the residual profit or loss and the second with a benchmarked RoA requires diligence and consistency but can be structured in a way that is well supported and designed to withstand tax audit scrutiny.
Digital hub service fee – if the contribution from the headquarters or digital hub is seen as a service and the functional profiles of the parties are structured as such, the possibility to establish a service fee remuneration for the centre is often simple to implement. However, because smart factory benefits are often shared across the network and depend on a central vision, it may be challenging to justify remunerating the centre solely through a benchmarked service fee.
Other models exist, and all require an understanding of how industrial data will be captured, managed, combined, used, and deployed. Operating choices need to be made with respect to the actual functional and risk profile of the associated entities involved in such a transformational project. Even if business choices come first, integrating transfer pricing considerations into the decision-making process can help ensure tax and transfer pricing compliance over the short, medium, and long term.
