High quality comparables are crucial when it comes to setting an arm’s length price in a transfer pricing (TP) analysis. A key factor in this process is making an informed choice about the dataset size, i.e. using comparable financials for one year or multiple years. This article explores key risks and factors to consider in setting an arm’s length price of transactions and using comparables for one or more years. We will be referring to the general rules of Latvian law and the TP guidelines issued by the OECD, with an example from case law.
Latvian law does not prescribe a position for or against using comparables for multiple years. Section 15.2(13)(1) of the Taxes and Duties Act provides that in preparing TP documentation, the taxpayer must use the most relevant information that is reasonably available within the financial year. Paragraph 3.2.9 of the Cabinet of Ministers’ Rule No. 802 explains that if data for multiple years is used in preparing the local file, the taxpayer must state the reason for taking this approach.
The OECD guidelines do not strictly provide for a single approach and do not state any particular number of years as a period for using data. The guidelines (from paragraph 3.75 onwards) explain that analysing data for multiple years is useful in practice, yet this is not a systematic requirement and the multiple-year approach should only be adopted where it adds value to the TP analysis.
The Supreme Court's Ruling SKA-95/2020 deals with a situation where differences of opinion between the State Revenue Service (SRS) and a taxpayer on using comparables for one year or multiple years resulted in an arm’s length range that varies considerably.
Let us examine this case where the taxpayer provided stevedore services to a related foreign company and used comparables for three years in the TP analysis, arriving at a markup of 5.40% on those services. The SRS rejected this approach and used comparables for only one year instead to arrive at a considerably higher arm’s length markup of 15.12%.
The taxpayer defended his approach by invoking paragraph 3.57 of the OECD guidelines, which suggests that statistical tools could make the analysis more reliable. In this case the taxpayer chose to analyse data for three years and claimed that an in-depth economic and market analysis places an unreasonably heavy administrative burden contrary to paragraph 3.69 of the guidelines. The taxpayer pointed out that so far the SRS had not challenged the use of data for three years and emphasised that their approach should be uniform and consistent.
The SRS claimed that the taxpayer’s approach fails to provide an objective view of arm’s length prices and repeatedly emphasised the significance of an economic and market analysis because it not only provides details of comparable transactions between the companies in previous years but also compares previous years and the year of the controlled transaction.
The SRS pointed out that in recent years the profit indicators of comparable companies have varied widely, leading to a wide range of profit margins (including by year) and that the indicators of certain companies are substantially affecting the three-year average operating profit margin used in calculating the profit indicators, which consequently fails to provide an objective view of the arm’s length price. The SRS recognised that the most accurate way of determining the taxpayer’s markup on operating costs is based on comparables for the last of the three years selected.
In its ruling the Supreme Court upheld the SRS approach, leading to the following conclusions:
Our recent communication with the SRS about using information for benchmarking studies highlights the most reliable information: the terms of comparable uncontrolled transactions taking place in the same period as the controlled transaction.
If you have any comments on this article please email them to lv_mindlink@pwc.com
Ask question
