Benchmark studies: criteria and data availability

Benchmark studies for comparable net margins ("BMs" – also referred to as database studies) are the cornerstone of numerous transfer pricing documentations. In particular, for foreign multinational enterprises with routine entities in Germany (e.g., contract manufacturing or low-risk distribution), a BM frequently serves as the primary basis for demonstrating compliance with the arm’s length principle (the same applies in outbound cases).

In short, a net-margin BM represents the typical practical application of the Transactional Net Margin Method ("TNMM"). By contrast, benchmark studies may also be prepared for comparable royalty rates, interest rates, franchise fees, etc.; these, however constitute applications of the Comparable Uncontrolled Price Method ("CUP"). The TNMM is one of the five commonly used transfer pricing methods and is arguably the most frequently applied method in practice. The remaining four methods are the Comparable Uncontrolled Price Method, the Cost Plus Method, the Resale Price Method, and the Profit Split Method. Under the TNMM, the net margins of independent comparable companies are generally compared with the net margin of the respective routine entity to assess whether its remuneration is consistent with the arm’s length principle. 

Our practical experience from tax audits in Germany shows that both the results and the set-up of benchmark studies (in particular the search criteria applied and the timeliness of the study) are often key areas of discussion. Unlike the Administrative Principles - Procedures (2005), which have since been repealed, the current Administrative Principles on Transfer Pricing (2024) are now aligned with the OECD Transfer Pricing Guidelines for Multinational Enterprises and Tax Administrations (2022). It is therefore all the more important that benchmark studies are prepared in accordance with the best-practice criteria prescribed by the OECD and recognized in Germany.

This newsletter article therefore examines which criteria for benchmark studies are recognized in Germany and when it is appropriate to update benchmark studies.

The OECD identifies a number of criteria that should be considered when conducting benchmark studies. The German tax authorities generally follow the OECD Transfer Pricing Guidelines, subject to the requirements of German tax law. In addition, further requirements and best practices have emerged over time, particularly as a result of extensive experience from tax audits, and regularly play an important role in the preparation and defense of benchmark studies. The overview below summarizes what we consider to be the key criteria for benchmark studies in Germany. It reflects both the OECD recommendations and important practical insights gained from many years of tax audit experience; however, it is not intended to be exhaustive. 

• Independence criterion: This criterion is essential to ensure that only independent companies are included in the benchmark study. For German transfer pricing purposes, it should be ensured that any shareholding in the comparable companies remains below 25%. The 25% threshold is derived from Section 1 (2) of the German Foreign Tax Act (Außensteuergesetz, “AStG”).
• Pan-regional search approach: German comparable companies are generally preferred. Due to the limited availability of suitable local comparables, a regional approach is frequently adopted, for example, a pan-European search strategy for German routine entities. This typically includes the EU-27 countries together with the United Kingdom, Norway, and Switzerland. This approach is internationally accepted and has been empirically validated (see Meenan et al. (2004), “Is Europe One Market? A Transfer Pricing Economic Analysis of Pan-European Comparables Sets”; Schwarz et al. (2018), “Die Vergleichbarkeit von Unternehmen bei pan-europäischen Benchmarkstudien”).
• Multi-year analysis: Most German tax audits cover a review period of three to five fiscal years. Accordingly, it is considered best practice to prepare benchmark studies on a multi-year period (e.g., for the period 2022–2024). Considering multiple fiscal years can enhance the reliability of a benchmark study, as it helps mitigate the effects of economic fluctuations, industry and product life cycles, and extraordinary one-off events in individual years. Accordingly, both the OECD and the EU Joint Transfer Pricing Forum (see EU JTPF (2016), “Report on the Use of Comparables in the EU”), whose recommendations are also taken into account by the German tax authorities, recommend the use of multi-year data. Under the EU JTPF recommendations, a multi-year analysis should generally cover a period of three to five years and include at least three fiscal years. 
• Revenue threshold: Minimum revenue thresholds should be determined on a case-by-case basis and primarily serve to exclude very small companies. As a general rule, it is not possible to set the revenue threshold at a high level (e.g., EUR 100 million), as this would reduce the available data pool too significantly. In addition, larger companies are often part of a group and are therefore not suitable for comparability purposes. We therefore recommend applying a minimum revenue threshold of EUR 1 million in order to exclude very small companies that are structurally not comparable. 
• Method for calculating the IQR: Various statistically recognized methods may be applied to calculate the interquartile range. Neither the OECD Transfer Pricing Guidelines nor the current Administrative Principles provide specific guidance on the calculation of quartiles. It is therefore essential that a recognized statistical method is applied consistently and that its application is documented in a transparent and comprehensible manner. 
• Minimum number of comparable companies: There is no statutory minimum number of comparable companies. Rather, the key consideration is that the final set of comparable companies should be as reliable and statistically robust as possible in order to ensure reliable results. Both the OECD Transfer Pricing Guidelines and administrative practice place greater emphasis on the quality and comparability of the selected companies than on their sheer number (“Quality over Quantity”). Accordingly, the required number of comparable companies will always depend on the specific facts and circumstances of the case as well as on the quality of the available comparable data. 

In addition to the criteria described above, it is essential to keep a benchmark study up to date for ongoing transactions. Under the OECD Transfer Pricing Guidelines, and therefore also under German administrative practice, a benchmark study should be renewed every three years. In the two years between full renewals, it is generally sufficient to perform a so-called financial update, i.e. to review, in addition to updating the financial data, the activity status and independence of the comparable companies, but not their functions. 

In connection with the update of benchmark studies discussed above, the question frequently arises in practice as to the most appropriate timing for such an update or the preparation of a new study. In other words, at what point are the required financial data available in sufficient quantity within the database? This aspect in particular can have a significant impact on the reliability of a benchmark study. The following analysis does not further examine the issue of data availability at the level of individual countries. Based on our experience, the composition of regional comparable sets generally changes only to a limited extent over time. Certain countries, particularly Italy, regularly account for a significant share of the comparable companies due to the greater availability of published company data, whereas German comparable companies are often represented much less frequently (see also the above discussion on the pan-regional search approach). 

To analyze the timing of data availability, we conducted an illustrative study based on the OrbisEurope database (“Orbis”) published by Moody’s. Orbis is the most widely used database in Europe and is also used by the German tax authorities. The period under review covers the years 2022 to 2024. The May 2026 database version was used as the reference for a complete data set. Applying an illustrative search strategy based on commonly used criteria, a total of 3,211 companies were identified for the analysis period from 2022 to 2024. For the purposes of our analysis, this data set represents data availability of 100%.

Subsequently, an analysis was performed to determine how many of these 3,211 companies already had complete financial data for the years 2022 to 2024 available in earlier database versions. As a logical starting point for the analysis, the January 2025 database version was selected (as it is unlikely that a company would have published its 2024 financial statements before January 2025). By examining the successive database versions (i.e. all database versions from January 2025 to May 2026), it is possible to trace how the availability of the financial data required for benchmark studies developed over time. 

The results reveal a significant time lag between the end of a fiscal year and the actual availability of reliable benchmark data. In January 2025, only 16 of the 3,211 companies that were ultimately available were included in the dataset (0.5%). Even by mid-year, in June 2025, data availability remained incomplete at only around 10%. The August database version reached data availability of 50%, which increased to approximately 80% with the October version. 

According to our analysis, a sufficient volume of data for conducting or updating a benchmark study is not available until the last quarter of the following year. The figure below illustrates the increase in data availability over the course of the year. It is also important to note that the respective database versions are only released around the middle of the following month. This means that the October version typically only becomes available in mid-November.

Figure: Cumulative Development of Available Financial Data in Orbis from January 2025 through May 2026 

Source: KPMG based on Moody’s OrbisEurope database. 

Foreign multinational enterprises with German entities (or German multinational enterprises) should ensure that the key criteria outlined above are taken into account when preparing their benchmark studies in order to reduce the risk of adjustments during tax audits. As a general rule, benchmark studies should be updated on an ongoing basis. Accordingly, a new benchmark study should be prepared every three years, while financial updates of the comparable companies will generally be sufficient in the intervening years. 

Another important aspect in the preparation of benchmark studies is data availability. Our analysis has shown that it regularly takes approximately one year for the financial data of almost all companies in the OrbisEurope database to become available. Around 80% of the financial data are available from mid-November of the following year. Based on these findings, the general recommendation is that updates of benchmark studies should ideally not be initiated or carried out before November.

_{(1) Applicability of Section 153(4) of the German Fiscal Code (AO) to Transfer Pricing Issues Arising from Mutual Agreement Procedures – Part 3}