Transfer Pricing and Its Acceptance in Switzerland: From Arm’s Length Tradition to OECD Methodology

Switzerland, with its dense network of over one hundred double taxation agreements and its longstanding commitment to the arm’s length principle, has progressively aligned its practice with OECD transfer pricing guidelines — culminating, since 2024, in the express endorsement by the Swiss Federal Tax Administration of the five OECD methods and the accompanying comparability analysis methodology. 

This article examines the key concepts underpinning transfer pricing — from the delineation of controlled transactions and the comparability analysis to the selection and application of pricing methods — and situates them within the Swiss legal and administrative landscape. 

What Is a Controlled Transaction? 

A controlled transaction is not limited to a formal legal arrangement between two separate entities. It encompasses any arrangement — including the absence of one — where a fair share of profit is insufficiently or not at all allocated among related parties. This notion becomes especially peculiar in the relationship between a company’s residence state and its permanent establishment: there is no transaction in the strict legal sense, as the same legal entity is involved on both sides, yet the arm’s length principle must still be respected under applicable double taxation agreements. 

“Even the absence of a deal can be a deal — and that is the whole point.” 

Group synergies deserve particular attention. Where deliberate and concerted actions generate profits or losses, these must be compensated accordingly. Mere incidental benefits arising from group membership, however, do not call for compensation. Location savings illustrate this well: such savings should be allocated among group members in proportion to their respective contributions. 

Business restructurings raise additional complexity. A restructuring typically comprises multiple transactions and fundamentally shifts the functional analysis landscape. There are winners and losers within the group, even if the restructuring benefits the group as a whole. The appropriate allocation of gains and losses must therefore be examined carefully, which explains why restructurings feature so prominently in transfer pricing compliance documentation. 

“Where there are winners, there must be losers — and transfer pricing insists on accounting for both.” 

Intra-group services also require careful scrutiny. Certain activities — such as pure shareholder activities or duplicative administrative functions — arise solely from belonging to the group and would not be incurred by independent companies. These do not require compensation. Genuine intra-group services, however — those that would otherwise have been procured from independent providers — must be adequately remunerated. 

Finally, the transactions themselves must be properly interpreted. While it is unacceptable to disregard a transaction merely because adequate comparables are absent, it may be appropriate to compare it against alternative arrangements that independent parties would have pursued to achieve the same goal. This can involve splitting a single transaction into several components, or conversely, combining multiple transactions and evaluating them as a whole. Cost Contribution Arrangements are a notable example: they establish a framework of joint cooperation toward a common objective — akin to a joint venture or a société simple under Swiss law — where participants share upsides and downsides, whether directly through monetary compensation or indirectly, for instance by developing a common brand. In such arrangements, it is essential to distinguish true participants from mere service providers and to evaluate the consequences of entry (buy-in), exit (buy-out) and termination. Another concern, particularly with intangibles, is so-called “smurfing,” where multiple intangibles transferred together hold a combined value greater than the sum of their parts, and the group artificially reduces total remuneration by splitting the global transaction into separate ones. 

Why Comparability Analysis Matters 

It is insufficient to compare transactions in isolation. Even where comparable goods or services exist, differences in functions, assets, risks and context may render a superficial comparison misleading. This is precisely why the Comparable Uncontrolled Price method, while straightforward in application, is highly sensitive to even minor differences in functional analysis. 

“A comparison is only as good as the context surrounding it.” 

The functional analysis requires assessing, for each group member involved, what functions are performed, which assets are contributed and which risks are assumed. 

Regarding assets, intangibles demand particular attention. They must be specifically identified — broad or vague references are not accepted. Mere synergies or non-unique know-how that a competitor could replicate are not considered intangibles. However, the concept of assembled workforce merits special consideration: the benefit of avoiding the cost and time of assembling a specialized team that works cohesively together, readily provided in a transaction, cannot simply be ignored. Intangibles must further be evaluated according to transaction type (ownership versus licence) and level of protection (patent versus unregistered intellectual property). Crucially, mere ownership or funding of an intangible does not in itself justify an allocation of profits from its exploitation. A DEMPE analysis is required to determine which parties perform significant functions of Development, Enhancement, Maintenance, Protection and Exploitation. 

“Owning an intangible is not enough. What matters is who develops, maintains, enhances, protects and exploits it.” 

As for risks, while it is generally true that greater risk warrants greater reward and that an independent party would accept lower reward for reduced risk, mere formal approval is irrelevant. What matters is whether a party has effective decision-making and execution power to accept or refuse risk, to prevent it, and to respond to it — this constitutes risk management. A party must also have the financial capacity to assume risk. Even where risk assessment, mitigation and response are delegated, the delegating party retains a degree of control through the choice of specialist, the instructions given and the evaluation of results — an approach analogous to cura in eligendo, instruendo, custodiendo. If a party bearing risk consequences lacks genuine risk management and risk assumption capabilities, the risk management activities must be duly remunerated and the risk assumption duly compensated. While no definitive categorization exists, risks are typically grouped into market and competition risks; infrastructure and operational risks; transactional risks (debtors, procurement); hazards (usually insured); and financial risks (liquidity, credit, etc.). 

The comparability analysis unfolds in several key dimensions. Legal agreements provide a useful starting point but must be interpreted with a substance-over-form approach, paying attention to negotiations, actual execution and subsequent modifications — particularly whether such modifications reflect genuine evolution or reveal the parties’ original intent inconsistent with what was stipulated. This substance-over-form reasoning is well established in Swiss tax practice, for instance in recharacterizing loans as simulated constructive dividends, or in refusing safe harbour interest on short-term cash pooling that does not qualify as a typical loan. 

Product and service comparability is another essential dimension, particularly under the CUP method, while other methods are less sensitive to minor differences. 

Economic circumstances — geographic region, regulation, competition, market level — likewise bear significantly on comparability. One cannot blindly compare wholesale with retail. Regulatory conditions should not themselves justify deviation from the arm’s length principle, since independent parties face them too; however, regulations applying specifically to a group, such as anti-monopoly rules not applicable to smaller independent entities, must be taken into account. Geographic segmentation matters, though data availability and quality vary across jurisdictions, sometimes making it appropriate to rely on broader global metrics. 

Business strategy is also critical. Market penetration strategies involving temporary losses and price dumping may be justified by expected future gains, so the comparability analysis should consider the entire prospective horizon, not merely the loss-making period. While independent companies also incur losses, they would not do so indefinitely. A group member permanently absorbing losses for the group’s benefit is an indicator that it is not remunerated at arm’s length. That said, a group member may sustain losses longer than an independent enterprise if the group’s coordinated market penetration strategy ultimately promises greater profits. 

“Losses can be strategic — but they cannot be permanent. That is where the arm’s length principle draws the line.” 

Transfer Pricing Methods 

Transfer pricing is not an exact science. Parties must select the most appropriate method, which may require examining more than one — but not necessarily all — in accordance with the principle of proportionality. What is essential is the justification of that selection, based on strengths and weaknesses of each method, available data (internal or external comparables), the chosen tested party and the degree of comparability. While the list of five endorsed methods is theoretically non-exhaustive, in practice, justifying any alternative would be extremely challenging. 

“Five methods, none perfect — and choosing the right one is itself the first test.” 

General experience shows that intangibles are typically evaluated under CUP (where functional differences from independent parties are negligible) or the Profit Split Method (where both parties make unique contributions). Financial transactions rely heavily on CUP due to abundant external comparables and broad functional similarity among financial service providers. Prevailing guidance also suggests that CUP should be preferred when applicable, and that traditional methods (CUP, RPM, CPM) should take precedence over transactional profit methods (TNMM, Profit Split) when feasible. 

The Comparable Uncontrolled Price Method (CUP) is the simplest to apply, comparing prices directly. Its simplicity is its advantage, but even minor product or service deviations can render it inappropriate, and it is highly sensitive to functional analysis differences. 

The Resale Price Method (RPM) takes the uncontrolled resale price as a starting point, deducts an appropriate gross margin, and arrives at the arm’s length purchase price. Less sensitive to product variations than CUP, it is most suitable for distribution activities where the reseller adds little value — no significant intangibles or additional services. If the tested party’s margin is adequate, the price is arm’s length for both the tested party and the related-party supplier. The appropriate gross margin may be derived from internal or external comparables. 

The Cost Plus Method (CPM) starts from costs, adds an appropriate mark-up, and arrives at the arm’s length sale price. It is suited to manufacturing, finished goods and routine services where costs can be readily allocated to revenues. Its main difficulty lies in determining relevant costs — distinguishing operational from non-operational costs, allocating historical costs — and in avoiding contamination of the cost base by controlled transactions themselves. 

In practice, differences in accounting presentation across jurisdictions often make gross margin determination difficult, which tends to favour the TNMM with its reliance on net margins from external comparables. It is also worth noting that a lump-sum 5% cost-plus mark-up is widely accepted — by both OECD guidelines and the Swiss Circular — for low-value support intra-group services, though this pragmatic exception must never be applied to services that constitute core business activities. 

“Five percent sounds simple. But knowing when it applies — and when it absolutely does not — is everything.” 

The Transactional Net Margin Method (TNMM) compares net profit indicators (NPIs) by establishing a net profit on the controlled transaction relative to a base (revenue, costs or assets), then benchmarking this ratio against internal or external comparables. Net margins are less sensitive to major functional differences, making TNMM suitable for complex activities with high-value contributions. It also works well for intermediary transactions where both input and output are controlled, since net margins are less affected by cost contamination — the Berry ratio is specifically applied in such cases. However, net margins have less direct relation to the transaction itself, and TNMM may only be applied to a less complex tested party, making it a one-sided method. Where both parties make unique and valuable contributions, the Profit Split Method is more appropriate. Common NPI bases include: net profit margin (return on sales) for low-value distribution; net cost plus (return on total costs) for services and manufacturing; return on assets for capital-intensive activities; and the Berry ratio for intermediary activities — the latter being reliable only where the tested party’s functional value is proportional to operating expenses, not materially driven by the value of goods distributed, and the tested party does not perform other significant functions. 

The Profit Split Method is the most complex, particularly because external comparables for contribution analysis are largely unavailable. It is most suitable where both parties make unique and valuable contributions, such as joint R&D. The method seeks to determine a reasonable allocation of profits or losses in proportion to functions performed, assets contributed and risks assumed. The process involves a contribution analysis followed by a residual analysis allocating results using other methods, typically TNMM or CPM. Benchmarks for contribution analysis may be asset-based, cost-based or derived from other variables such as incremental sales, headcount or time. While the most subjective of the methods, it is the only appropriate measure where both parties bring unique and valuable contributions. 

In both the choice of method and its application, differences in products, services or functional analysis must either have no material effect on the price, or reasonable adjustments must be made to eliminate them. Blind transposition of benchmarks without adjustment is rarely appropriate. Extreme results — unusual profitability or losses — require heightened scrutiny. Nevertheless, because transfer pricing is not an exact science, deviations are tolerated: a price falling within the interquartile range of comparables is accepted as arm’s length. The OECD clearly warns countries against marginal adjustments. 

The quality and reliability of comparables are paramount. Internal comparables are inherently more appropriate than external ones. Time period and region must be considered in their selection, and the OECD has pushed further with specific guidance on intangible valuations. Importantly, tax authorities often have access to confidential information unavailable to taxpayers. Switzerland has expressly prohibited administrations from using such data against the taxpayer, and the constitutional principle of good faith further supports this — including through the business judgment rule, since the situation ex post is always clearer than it was ex ante. The OECD acknowledges that administrations tend to evaluate hard-to-value intangibles using ex post data, and endorses that such retrospective assessment should not take place where parties were genuinely unaware of relevant factors at the time, their ex ante evaluation approach was reasonable, or the deviation is minor (within 20%). While the OECD endorses flexibility, it does not accept simple rules of thumb (such as a 25/75 split between licensor and licensee) nor just any valuation approach — Discounted Cash Flow may be appropriate, but the discount rate must be reasonable, requiring sensitivity analysis to test how varying assumptions affect the outcome. In all cases, higher leniency in methodology demands higher reporting and transparency standards. 

“More freedom in method means more discipline in disclosure. The OECD gives with one hand and takes with the other.” 

Summary of the Comparability and Method Selection Process 

The sequence of comparability analysis, method selection and application can generally be distilled into nine key steps: timing; broad context; delineation of the controlled transaction and functional analysis; identification of internal comparables; identification of external comparables; choice of transfer pricing method; revision of external comparables to fit the selected method; adjustments; and conclusions. These steps are not strictly linear — any adjustment may require revisiting earlier stages. For instance, data examined under a chosen method may prove insufficient, prompting selection of another method and a fresh search for comparables. The comparability analysis thus operates in two phases: when delineating the controlled transaction, and when comparing it against independent parties to assess the arm’s length price. 

Acceptance of OECD Transfer Pricing Methodology in Switzerland 

The absence of group taxation — each company being treated individually — and the arm’s length concept are longstanding principles of Swiss corporate tax law. The legal bases suffice for primary adjustments (increasing profits from insufficient revenues or excessive expenses) and secondary adjustments (withholding tax on constructive dividends and, in fewer cases, issuance stamp tax on hidden capital contributions). The SFTA further recognizes that objective international profit allocation should apply between a permanent establishment and the corporate seat or place of effective management, even absent a double taxation agreement. Article 9 of the OECD Model Convention thus has a largely declarative character in the Swiss context. 

“Switzerland did not need the OECD to discover the arm’s length principle. It needed the OECD to apply it.” 

Swiss law does not contain an express statutory provision on transfer pricing. Rather, a long-established administrative practice and body of case law require respect for the arm’s length principle, grounded in the double economic taxation for corporations and the concept of constructive dividends embedded in both withholding tax and income tax legislation. 

The question was never whether arm’s length applies in Switzerland — it always has. The question was how it is evaluated. Historically, Switzerland gradually abandoned lump-sum and simplified practices, referring more and more to OECD guidelines in its circulars, while the Federal Court has likewise relied on OECD transfer pricing guidelines. Since 2024, the SFTA has expressly endorsed the five OECD methods and the comparability analysis methodology. 

The safe harbour rules on hidden capital and interest rates remain applicable and are not prohibited by the OECD. They play an important role in burden-of-proof allocation: the theory of constructive dividends in Switzerland requires that the unjustified advantage be recognizable, which it is when safe harbour rules are not respected. Non-compliance also reverses the burden of proof, creating a rebuttable presumption that the arm’s length principle has been breached. The taxpayer retains full capacity to demonstrate compliance using OECD methods, but the moment the taxpayer deviates from safe harbour rules, the administration is no longer bound by them either. In any event, safe harbour rules are insufficient to cover the full spectrum of products and services, which inevitably must be assessed through the lens of OECD transfer pricing guidelines. 

“Safe harbours protect — but only up to a point. Step outside them, and the burden shifts.” 

Swiss acceptance extends beyond methodology to certain procedural implications. While the OECD recognizes procedure as a state prerogative, various OECD guidelines within the scope of BEPS and double taxation agreements are broadly applicable in Switzerland. 

Unilateral advance pricing agreements already exist as a mechanism and are applicable to transfer pricing. The OECD prefers bilateral or multilateral APAs involving multiple states, and Switzerland permits this in the international context through the State Secretariat for International Finance (SFI) rather than the SFTA. 

The Mutual Agreement Procedure, strongly encouraged by the OECD in transfer pricing matters, is also available in Switzerland and may be initiated by the taxpayer through SFI, although the taxpayer does not become a formal party to the procedure. MAPs further allow avoidance of secondary adjustments — for instance, refraining from levying withholding tax on outbound constructive dividends where funds are repatriated to Switzerland, or conversely, refusing to levy withholding tax on compensatory payments from a Swiss entity to a foreign entity following a foreign primary adjustment, which would otherwise be recharacterized as a constructive dividend themselves. Switzerland would likewise apply a correlative adjustment through MAP, though not automatically — it must still be satisfied with the foreign state’s recalculation. 

International cooperation represents another significant dimension. APA rulings are often exchanged through spontaneous information exchange, with Switzerland acting both as sender and recipient. One non-negotiable OECD obligation — as opposed to the merely recommended Master File and Local File documentation — is Country-by-Country Reporting (CbCR), which Switzerland applies to groups exceeding CHF 900 million in consolidated turnover. Such groups either submit the report in Switzerland or Switzerland receives it from partner states. Even absent an ultimate parent company in Switzerland, other group entities with an office or permanent establishment may be compelled to submit the report on the group’s behalf, where the structural and geographical organization is such that Switzerland would not otherwise receive a CbCR from abroad. 

Conclusion 

What is clear about both the Swiss and OECD approaches is that transfer pricing practice provides a flexible yet firm framework for taxpayers and authorities alike — compelling the former to justify and transparently report, and the latter to levy and control taxation in good faith. 

“Flexibility for the method. Rigour for the reporting. That is the bargain at the heart of modern transfer pricing.”