Transfer pricing and the inherent arm’s length principle has been around for some time now and many African countries have implemented transfer pricing regulations. In Africa, only about five countries had transfer pricing legislation in place before 2000, now in 2019, only about five countries do not have any arm’s length or anti-avoidance provisions in place dealing with transfer pricing. Out of these countries approximately 40 countries in Africa have some sort of formal transfer pricing documentation requirements, be it in the form of formal submission or retention requirements.
Many articles, blog posts or formal discussion documents from authors around the world have dealt with transfer pricing in different forms. It seems we are reaching some fatigue around the theoretical application of transfer pricing and as such I thought I would bring some practical guidance to many issues faced within the South African transfer pricing space. Each paragraph below touches on key practical transfer pricing issues identified and a potential solution on how to approach them.
Taxpayers are awaiting further guidance from SARS in relation to transfer pricing. We have some guidance through Practice Notes and Interpretation Notes. More specifically, there is Practice Note 2, which was repealed through a draft Interpretation Note but was also made redundant through changes to the transfer pricing legislation (section 31). Furthermore, there is Practice Note 7 with certain amendments and there is also the infamous thin cap draft Interpretation Note. The issue with these notes is that they are outdated, repealed/redundant or in draft. SARS has not yet provided updated guidance for the latest version of section 31 but Practice Note 7 (which was published in 1999) and the South African legislation are inspired by the OECD Guidelines. As such, when in doubt and not sure how to treat something from a transfer pricing perspective, it seems reasonable to use the OECD Guidelines for guidance unless there are specific local legislative rules that contradict this or reservations by SARS on certain elements of the OECD Guidelines. For example, the low value-adding section in the OECD Guidelines comes to mind. SARS has mentioned through different platforms that it does not follow this, as a detailed analysis of all service transactions is required, including a benefit test.
Benchmarking studies or comparability analyses seem to raise many questions, with the key questions being around what comparable data is acceptable. Theoretically, comparables in the same industry, geographical region with the same functional profile should be considered, but we are all aware it is an impossible task to get reliable data that meets these standards. It is probably no secret that global benchmarking studies are frowned upon by most African tax authorities and if no local comparable companies exist, it isn’t always clear how to approach this dilemma. The following points may assist in getting to a solution when faced with having no comparable data and data that could be seen non-comparable. When looking into comparable countries or regions, it may be useful to select similar countries based on country credit ratings to that of the tested party’s country. Additionally, one can also look into geographical regions such as the BRICS to which South Africa belongs. There may be a stage in a benchmarking study where one has to choose between functional or geographical comparison, in this case functional comparison seems more important as there are ways to adjust for geographical differences, but none really for functional differences with the exception of working capital adjustments. As per the OECD Guidelines, certain adjustments to comparable data is acceptable as long as this will result in more comparable data.
The above discussion around comparable data raises the question of what is an arm’s length range. Many tax jurisdictions default to an interquartile range (IQR) but bear in mind, where each observation is as comparable as the other, arguably the full range could be seen to be an arm’s length range. This is unlikely to happen in a benchmarking context but may well be the case in some comparable uncontrolled price (CUP) studies. Even though our focus is on South Africa, it is worth mentioning that some African countries narrowed the range in relation to an arm’s length range. For example in Tanzania the arm’s length range is considered to be the 35th percentile to 65th percentile unlike the IQR which is from the 25th percentile to the 75th.
A last point worth making in relation to benchmarking studies is that, where the functional profile of the tested party as well as the comparables hasn’t changed materially, a benchmarking study can be updated by performing financial updates to the comparable data. This is in line with the OECD Guidelines and a benchmarking study can be updated twice (or for two additional years), after which a new benchmarking study should be performed.
For the last part of this article, I wanted to touch base on the submission of transfer pricing documentation, specifically for taxpayers that only have to submit the master file and local file. Where the threshold is met, a submission must be made through SARS’ e-filing portal. It seems not everyone is aware of this, but the transfer pricing documentation must be submitted via the CBC tab in the portal and not as supporting documentation with the ITR14. The issue with submitting the documentation as supporting documentation and not via the CBC tab is that different teams within SARS are reviewing the different tabs. The CBC tab is not a default and will have to be activated before it is visible on e-filing. For those who have not submitted the master file and local file via the CBC tab it is recommended to submit this correctly as soon as possible. The good news is that the current legislation does not yet provide for penalties on the wrong submission of the master file and local file but only touches on penalties for the country by country report. Furthermore, when submitting via the CBC tab, taxpayers will receive a confirmation that the documentation has been uploaded and SARS will also provide feedback if there are gaps in the uploaded documentation.