Base erosion and profit shifting – a South African perspective

OECD image 1Author: Peter Dachs – ENSafrica

The concept of base erosion and profit shifting (BEPS) has been debated at various international forums following discussions at the G20 Finance Ministers and Central Bank Governors meeting and the G20 Heads of State summit in Russia last year.

The Organisation for Economic Co-operation and Development’s (OECD) BEPS Action Plan provides for 15 actions to be completed in three phases by December 2015.

One of the first of these is an in-depth report identifying tax challenges raised by the digital economy and the necessary actions to address them. The report includes recommendations on the design of domestic and tax treaty measures to neutralise the effects of hybrid mismatch arrangements both from a domestic and treaty law perspective. It also makes recommendations on the design of domestic and tax treaty measures to prevent abusive tax treaties as well as changes to transfer pricing rules in relation to intangible assets.

The Davis Tax Committee has been tasked with addressing BEPS in a South African context, which includes hybrid mismatch arrangements.

According to the OECD, hybrid mismatch arrangements exploit the differences in the tax treatment of instruments, entities or transfers between countries. These arrangements generally use one or more of the following elements:

  • Hybrid entities that are treated as transparent for tax purposes in one country and non-transparent in another;
  • Dual-resident entities, which are resident in two different countries for tax purposes;
  • Hybrid instruments that are treated differently for tax purposes in the countries involved, for example as debt in one country and equity in another; and
  • Hybrid transfers, which are arrangements treated as transfer of ownership of an asset in one country, but as a collateralised loan in another.

Hybrid mismatch arrangements include double-deduction schemes, where a deduction related to the same contractual obligation is claimed in two different countries; schemes that create a deduction in one country, but avoid the corresponding income inclusion in another country; and arrangements that generate foreign tax credits that would otherwise not be available .

Among concerns noted by the OECD are:

  • While it can be difficult to determine which of the countries has lost tax revenue, it is clear that the countries concerned collectively lose tax revenue.
  • Businesses that use mismatch opportunities have competitive advantages over those that cannot.
  • Where a hybrid mismatch is available, a cross-border investment will often be more attractive than an equivalent domestic investment.
  • The adoption of tax-driven structures leads to a lack of transparency, with most people generally unaware that the tax regime is quite different for those using mismatch opportunities.
  • Mismatch opportunities are more readily available for taxpayers with income from capital, rather than labour.

From a South African tax perspective, transactions that involve hybrid entities or instruments may result in the claiming of foreign tax credits in circumstances where the foreign tax suffered is effectively neutralised in the foreign jurisdiction.

Alternatively, such arrangements result in exemption from tax in respect of foreign-sourced income by virtue of an appropriate double-tax agreement.

Repurchase transactions are common hybrid entity arrangements. There are several variations of these transactions, but the key mechanics are essentially the same:

  • A partnership is set up in Jurisdiction A;
  • An investor in Jurisdiction B acquires an interest in the partnership in terms of a repurchase agreement. The investor may borrow money to acquire this “partnership interest”;
  • The partnership uses its capital to invest in a loan and earns interest;
  • The investor is, in terms of Jurisdiction B’s tax laws, entitled to its share of the partnership income derived in accordance with the partnership agreement.

From Jurisdiction A’s tax perspective, the partnership is viewed as a separate entity and is liable to tax. The partnership therefore pays tax and the investor claims a credit for the tax suffered in respect of its partnership distributions. The partnership then distributes a post-tax return to the investor.

The investor is entitled to a credit for the tax paid by the partnership. The investor may also claim a deduction for its funding costs to acquire the partnership interest.

However, Jurisdiction A provides a tax credit to the seller of the partnership interest in respect of the amount of tax paid by the partnership. The tax paid by the partnership is therefore neutralised since, on a consolidated basis, no tax is suffered in Jurisdiction A.

By way of an example of dual-resident entities and hybrid instruments, a company is incorporated and tax resident in Jurisdiction B. The company issues redeemable preference shares to investors in Jurisdiction A and invests in debt instruments issued by companies in Jurisdiction A.

From Jurisdiction A’s tax perspective, the company earns interest income that is not taxed in Jurisdiction A due to the provisions of the double-tax agreement between Jurisdiction A and Jurisdiction B.

The company pays out dividends that are tax exempt in the hands of the investors in Jurisdiction A.

From Jurisdiction B’s tax perspective, the interest received by the company is taxable. However, the redeemable preference shares are re-characterised as debt for tax purposes and therefore a deduction is granted for the dividends paid on these shares. The company is therefore only taxed on a spread/margin in Jurisdiction B.

The issue arises whether transactions such as these erode the South African tax base and, if so, how they have been dealt with in terms of South African law.

We do not have any case law in respect of foreign tax credit transactions. However, various New Zealand cases have dealt with foreign tax credits and disallowed such credits in the hands of the New Zealand entities. Although only of persuasive influence in South Africa, the New Zealand courts held that the foreign tax credits could not be claimed and set out detailed reasoning in this regard.

It also needs asking whether there is any policy or principle issue with a tax credit being granted in circumstances where foreign tax is suffered by a foreign entity, but where such tax is then effectively neutralised in the foreign jurisdiction.

It may be argued that to place an onus on a South African taxpayer to prove that the foreign tax was not “economically neutralised” in some manner is to place too high a burden on such taxpayers.

In respect of current law, section 6quat of the Income Tax Act and the provisions relating to the elimination of double taxation in double-tax agreements require that tax is paid in the foreign jurisdiction and that this represents a final tax by the relevant entity. It may be argued that this should be sufficient for a South African taxpayer to claim a credit in respect of foreign taxes.

South Africa has a general anti-tax-avoidance provision which is contained in sections 80A-L of the Act and may be used in appropriate circumstances. The first issue is whether a tax benefit exists as a consequence of the transaction.

In the context of the definition of a “tax benefit” in terms of section 1 of the Act, based on case law (Hicklin v SIR 41 SATC 179 and Smith v Commissioner for Inland Revenue 26 SATC 1), the liability for the payment of any tax, levy or duty that a taxpayer must seek to avoid, postpone or reduce is not an accrued or existing liability, but an anticipated liability. It was held that to avoid liability in this sense is “to get out of the way of, escape or prevent an anticipated liability”.

In ITC 1625, 59 SATC 383, it was held that the test to be applied in determining whether a transaction had the effect of avoiding tax was to ask whether “the taxpayer would have suffered tax but for the transaction”.  The court stated that “if the transaction in issue had not been entered into the taxpayer would not have acquired the property, it would not have earned the income and it would not have incurred the interest expenditure” and thus the court could find “no basis on which it can successfully be argued that by incurring expenditure on interest in order to earn the income on which it has to pay tax the taxpayer avoided tax or reduced tax”.

If there is a tax benefit, the second requirement for the application of the anti-tax-avoidance provisions is that the “sole or main purpose” is to obtain such benefit. Therefore, provided the taxpayer does not comply with this requirement, the arrangement will not constitute an impermissible tax avoidance arrangement.

If a South African resident invests in a transaction such as those described above instead of, for example, advancing a loan to an entity, then this will have the effect of sidestepping an anticipated tax liability in respect of the interest that the South African resident would have received. A “tax benefit” will therefore arise for the South African resident who will then bear the onus of proving that its sole or main purpose was not to achieve such benefit.

Without satisfying this requirement, the issue of base erosion can be seen not to apply since, in terms of the provisions of section 80B of the Act, the transaction may be re-characterised in a manner such that the South African resident is deemed to have advanced a loan to the ultimate borrower and received interest from this.

  • These are the main issues involved in hybrid mismatch arrangements, and other aspects of BEPS in a South African context will be discussed in future articles.