Tax Avoidance – Connected persons

In tax, as in business, it matters a lot whether or not you are connected. In most tax systems across the world, a taxpayer’s relation to other entities with which it transacts determines whether a range of anti-avoidance legislation will apply or not. Whether the parties are called “related parties”, or “connected persons” as is the case in South Africa, this implies a non-arm’s-length commercial relationship between transacting entities and forms the cornerstone upon which revenue authorities seek to justify the application of anti-avoidance legislation.

In the South African domestic tax law, transfer pricing provisions are applied to adjust prices in respect of transactions between resident and non-resident connected persons. Deemed market value proceeds apply in respect of the disposal of assets between connected persons and a deemed dividend with resultant Secondary Tax on Companies (STC) is triggered if any benefit is granted to a connected person in relation to the shareholder. Connected persons provisions also feature in the context of hybrid equity instruments, hybrid debt instruments and diversionary rules related to the attributable net income of controlled foreign companies.

With this substantial volume of anti-avoidance legislation that could be triggered if one enters into dealings with a connected person, it is essential to establish whether the transacting parties are connected persons when entering into transactions. South African taxpayers face a treacherous task of navigating extremely complicated, clumsy and disjointed legislation to determine their tax fate. Internationally, best practice requires legislation, and in particular anti-avoidance legislation, to be clear and understandable. In this, South Africa is unfortunately still lacking.

Taking the definition of a connected person as set out in section 1 of the Income Tax Act, Act  No. 58 of 1962 (the Act), a company is a connected person in relation to another company that would form part of the same group of companies if the expression “at least 70%” with reference to shareholding in the definition of a “group of companies” was replaced by the expression “more than 50%”. To put this in context, the definition of a “group of companies” also has a different meaning depending with which part of the Act one works. The first step is to determine the appropriate definition of a “group of companies”, and then apply the 50% exception to that definition to determine whether or not one complies with the first test of the “connected person” definition.

The second test for a company in relation to another company is that company will be a connected person if at least 20% of the equity shares in the company are held by that other company, and no shareholder holds the majority voting rights in the organisation. This seems straightforward to determine but, again, there are exceptions. For transfer pricing purposes, transactions relating to intellectual property and knowledge, the phrase “and no shareholder holds the majority voting rights in the company”, should be disregarded. This has the effect of lowering the threshold of a connected person for purposes of specified transfer pricing provisions. With effect from 1 October 2011 this lower threshold, and the required disregarding of the phrase, will be applied to all transactions, thus expanding the potential application of transfer pricing provisions.

These two tests apply in circumstances where there is a direct shareholding. Where there is no direct shareholding, a company will be a connected person in relation to another entity if such other entity is managed or controlled (more than 50% shareholding) by any person who or which is a connected person in relation to such company. In practice this requires careful analysis.

The final twist in this complicated situation is that the connected person test must be applied in converse. Therefore, if company B is a connected person in respect of company A, company A will automatically be a connected person in relation to company B. Sounds simple and logical, but quite different and difficult to test in complicated corporate transactions.

Let’s illustrate the application by a simple example: company A (A) holds 60% of the shares in company B (B) and 40% of the shares in company C (C). The other 60% of the shares in C are held by one shareholder. We need to establish whether C is a connected person in relation to B. For purposes other than transfer pricing, B will be a connected person in relation to A, as A holds more than 50% of the shares in B. C will not be a connected person in relation to A as, despite A holding more than 20% of the shares in C, the other shareholder in C holds the majority. As A does not control C, C is not connected to B, and B is also not connected to C.

For transfer pricing purposes, intellectual property and knowledge related transactions, and post

1 October 2011 transactions, the analysis is different. A will be connected to B as it holds more than 50% of the shares in B. A will also be connected to C as it holds more than 20% of the shares in C and the requirement for the lack of majority voting rights must be disregarded. C will not be a connected person in relation to B on account of the joint control provision, as A does not control C. However, this must also be tested conversely. As A is connected to C, and also controls more than 50%  of B, B is a connected person in relation to C on account of the joint control provisions. As B is a connected person in relation to C, C will also be a connected person in relation to B.

Considering that this complicated test is required for testing a simple example, it is of concern whether taxpayers are always aware that they are at risk of being subjected to anti-avoidance legislation. This is clearly not conducive to establishing tax certainty in South Africa and illustrates the need to overhaul complicated and clumsy tax legislation.