Illustration: KAREN MOOLMAN
Illustration: KAREN MOOLMAN

There are frequently reports of billions of rand leaving the SA tax net due to transfer pricing abuses. The narrative is broadly that SA multinational companies are shipping profits abroad to related parties situated in foreign jurisdictions (often tax havens) in contravention of the country’s transfer pricing rules.

In terms of SA’s transfer pricing rules, where related parties in different jurisdictions provide intellectual property, goods, services or finance to each other they need to comply with the arm’s-length test. This means the pricing of such products should be comparable to what the party would charge an unrelated entity.

The Organisation for Economic Co-operation and Development (OECD) has, over the years, come up with five methodologies to test whether transactions between related parties comply with the arm’s-length test. These methods are accepted and used by all OECD member states, as well as many nonmember states such as SA.

Although there is no fixed hierarchy of methods, there is a general ranking based on the nature of the transactions between the parties and the availability of the financial information needed to apply the arm’s-length principle.

For example, the most accurate way of determining an arm’s-length price is generally by applying the so-called comparable uncontrolled price method. This analyses the pricing of comparable goods between comparable entities in the relevant jurisdictions. So, if a multinational entity sells widgets to a related party in Spain and financial information is found on a comparable entity that also sells widgets to a comparable third party located in Spain, then this seems like a reasonable proxy for determining an arm’s-length price.

Unfortunately, finding the necessary comparability is easier said than done. What if the comparable entity was not based in Spain, but rather Portugal? Does this affect the nature of the market and therefore the application of the arm’s-length test? What if the widgets were almost, but not completely, identical?

So, while the international tax world recognises that the arm’s-length tests and methods prescribed by the OECD are imperfect, they are well established and widely understood and applied. SA’s reliance on these methods is supported by the OECD and its member states.

Reports alleging transfer pricing abuses in SA and attempting to put numbers to such abuse rely on various assumptions relating to the application of the arm’s-length principle and may also ignore the established transfer pricing methodologies set down by the OECD. One recent report compared prices on a basket of goods sold to SA taxpayers from entities in low-tax jurisdictions to the prices on a basket of goods sold from entities in high-tax jurisdictions.

But this does not provide accurate evidence that transfer pricing abuses have occurred. Some of the questions that arise are whether the goods are identical, whether they originated from the jurisdictions that sold them to SA and whether they were sold to related or unrelated parties.

In addition, transfer pricing rules are agnostic as to whether the related party is located in a tax haven or a high-tax jurisdiction. As long as that entity performs the relevant functions and assumes the risks required to justify its pricing, it complies with the transfer pricing rules. So, while these reports are useful in pointing out the existence of potential transfer pricing abuse, it is not easy to quantify the extent of any such abuses.

An advantage SA has over many other countries in combating transfer pricing abuses is its exchange control regulations. For example, if an SA entity wished to transfer its intellectual property to a subsidiary located in a tax haven, it would require consent for such transfer. This consent would not be forthcoming since the sale is to a related party. Instead, the exchange control authorities would require the intellectual property to be licensed to the offshore entity, thus ensuring a taxable royalty flow back to SA.

It should also be noted that since  2016 SA taxpayers  have been required to comply with stringent documentation requirements to satisfy SA’s transfer pricing rules. At a minimum, such taxpayers are required to prepare a so-called “local file” for submission to the SA Revenue Service (Sars). The local file must have detailed information and support on all the relevant intercompany transactions entered into with foreign related entities. This requirement arises from SA’s compliance with Action 13 of the base erosion and profit shifting report concluded by the OECD.

In addition, Sars has information available to it in relation to these transactions such as customs returns, exchange control applications, financial statements  and other correspondence with Sars.

SA companies that do not comply with the transfer pricing rules will have an adjustment made to their taxable income, resulting in income tax being imposed at a rate of 28%. In addition, dividends tax at a rate of 20% may be levied on such a transfer pricing adjustment. In terms of the Tax Administration Act of 2011, penalties of up to 200% may also be levied as well as interest on the unpaid tax.

In addition, the rules relating to prescription that generally mean Sars cannot raise any additional assessment after a period of three years from the date of the original assessment do not apply in circumstances where there has been fraud, misrepresentation or nondisclosure of material facts. It is likely that in the case of deliberate transfer pricing abuses, or simply in circumstances where the relevant transfer pricing methodologies have not been properly implemented, there will be nondisclosure of material facts by the taxpayer.

In these circumstances, the taxpayer will not be protected by the three-year prescription rule. Multinational entities that embark on transfer pricing abuses in SA would therefore face severe fiscal consequences as a result of such abuses.

• Dachs is head of ENSafrica’s tax department. He writes in his personal capacity