When discussing benchmarking in Transfer Pricing, the conversation often focuses on the result: the arm’s-length range and, where applicable, the median or another measure of central tendency used to evaluate a transaction. However, the preliminary process that determines how that final set of comparables is arrived at is rarely explained.
In practice, this process involves applying a series of quantitative and qualitative criteria to an initial universe—which may consist of hundreds or thousands of companies—until those that demonstrate a reasonable level of comparability and can withstand scrutiny by a tax authority are identified.
Quantitative Filters: The First Stage in Selecting Comparables
The process typically begins with a broad search of commercial databases, using criteria related to the economic activity of the entity under analysis, such as ISIC codes, NAICS codes, or equivalent classifications.
Objective quantitative criteria can be applied to this initial universe to identify companies that are potentially less comparable. Among other factors, the company’s size may be evaluated based on its revenue or assets, atypical financial situations, the presence of significant intangible assets, certain inventory levels, or financial indicators relevant to the operation under analysis.
The existence of significant transactions with related parties may also be evaluated to preserve the comparability of the company, by applying the parameters defined in the search methodology and considering the available information.
In the case of companies with recurring losses, they should not be automatically excluded. The OECD Guidelines state that it is necessary to analyze the circumstances underlying such losses and determine whether they result from normal business conditions or whether they indicate relevant differences in functions, assets, risks, or economic circumstances that could affect comparability.
Qualitative filters: where technical judgment comes into play
Once the quantitative stage is complete, the process requires a qualitative review on a case-by-case basis. It is at this stage that rigorous benchmarking truly differs from that based solely on automated filters.
This involves reviewing the business description of each candidate company based on its financial statements, annual reports, corporate websites, public records, or other available sources, with the aim of confirming that it indeed engages in activities comparable to those of the entity under analysis.
Sharing the same industrial classification code does not, in and of itself, guarantee comparability.
For example, a company classified as an electronics distributor might, in practice, also engage in manufacturing activities, possess significant intangible assets, perform high-value-added commercial functions, or have a risk profile that is significantly different from that of the entity under analysis.
These differences can typically only be identified through a detailed qualitative review of the business and not through automated database filters.
Why is incorrectly discarding a comparable as risky as accepting an inappropriate one?
One of the common errors in Transfer Pricing studies is applying quantitative filters mechanically, without adequately documenting the reasons behind the inclusion or exclusion of each candidate company.
This can weaken the study in two ways.
If the tax authority identifies a potentially comparable company that was excluded without sufficient technical justification, it may question whether the final set was constructed objectively or whether the selection criteria were used to obtain a specific result.
Similarly, including a company that passed the quantitative filters but, after a more in-depth qualitative analysis, exhibits significant differences in its functions, assets, risks, products, or business strategy can reduce the reliability of the arm’s-length range obtained.
Therefore, selecting comparables is not simply a matter of narrowing down a list of companies using automatic filters. It requires assessing whether the companies remaining in the sample have characteristics sufficiently comparable to the operation under analysis.
It is not enough to present the final comparables: the exclusion process must also be documented
A defensible benchmarking study should not be limited to presenting the final list of accepted comparable companies.
It must also transparently document the search criteria used, the databases consulted, the filters applied, the number of companies identified at each stage of the process, and the specific reasons justifying the exclusion of candidates that did not make it into the final set.
This traceability, which can be organized using an acceptance and exclusion matrix, makes it possible to demonstrate that the selection of comparables was based on objective and consistent criteria.
The OECD Guidelines emphasize that the process of identifying comparables must be transparent, systematic, and verifiable. In this regard, the ability to explain why a company was included or excluded can be just as important as the full-competition range obtained at the end of the analysis.
At TPC Group, we build our benchmarking processes by documenting each quantitative and qualitative criterion applied in a traceable manner, with the goal of ensuring that the final set of comparables can withstand both rigorous technical analysis and scrutiny by a tax authority.
Source: