Tax Avoidance and section 103 of the income tax act

Discussion Paper

February 2006.

  1. Background
  2. The South African Revenue Service (SARS) has issued a discussion paper on tax avoidance and ways in which tax avoidance can be addressed. SARS have invited comment from the general public on the paper and their proposals to amend section 103 of the Income Tax Act.

    In preparing this document, The Banking Association has reviewed the comments made by BUSA/SACOB, SAICA and KPMG.

    SARS are to be commended for the frank and open manner in which the problems of tax avoidance are identified and analysed. The document produced is courageous in its attempt to objectively examine the concepts of avoidance and evasion. It touches on some of the fundamental issues of corporate behaviour and ethics and exposes the role of tax planning in economic activity. Taxpayers have the right, within the provisions of the Income Tax Act to arrange their affairs in such manner that they do not incur a higher tax charge than is provided for in the Act and was intended by the legislators. Conversely, taxpayers have an obligation to pay those taxes as are legally required by the Act. It is the balance between these two approaches that has the potential to become difficult to manage in a fair and reasonable way.

  3. Financial Service Providers
  4. Many of the examples given in the discussion paper are illustrated with reference to the participation of banks in the arrangements. One of the main functions of banks is intermediation between borrowers and investors to mobilise funds for productive purposes. Inevitably, funds borrowed are repaid and deposits made are returned. At a very basic level, the circular flow of funds caused by borrowing and repayment cannot, in itself be indicative of anything other than normal banking practice. Banks facilitate the mobilisation of funds from those that deposit it to those that borrow it.

    In the intermediation process the costs of borrowing and the benefits of lending are factored into the margins earned by banks. Where allowances under the Act are available, these are availed of in the intermediation process. Where, for example, leasing allowances are intended by the legislature to stimulate capital investment, banks will give effect to that intention by claiming the allowances and passing the benefit on to the lessee in the form of lower rentals. Competition between banks for business will ensure that the intended beneficiary (the lessee) will benefit as intended by the legislature. Unless banks perform the intermediation role, many of the intended beneficiaries will not receive the stimulation intended. The intermediation role is therefore not limited to the flow of funds, but extends to income tax allowances. This process should not be seen to be undesirable in any way, as it is an essential part of fuelling the growth of the economy, by facilitating both local and foreign investment, which is vital in a developing economy.

    The Banking Association is concerned that the factors listed in the proposed section 103(2) are seen in many respects to be specifically aimed at the banking sector. For example, the existence of the rebuttable presumption contained in the proposed section 103(4)(b) could be extremely prejudicial to the ability of the banking sector to provide cost-effective finance, as there are numerous items listed in the proposed section 103(2), which occur in a normal manner in the industry on a day to day basis.

  5. Application of section 103
  6. In most tax disputes, SARS would normally have to prove a case against the taxpayer, to a greater or lesser extent. In contrast, the draft wording is heavily weighted in favour of SARS regarding presumptions, albeit rebuttable in theory. The practical effect, however, will be that it will be almost impossible to rebut many of the presumptions, thus making the civil procedure unfair and biased against the taxpayer.

    The proposed new section includes two rebuttable presumptions:

    It is unlikely that the taxpayer will ever be able to discharge the first presumption, in view of the fact that, in terms of section (3), this is to be determined objectively. The existence of a tax benefit will ordinarily be deemed to be one of the main purposes of the transaction.

    It will be extremely difficult for the taxpayer to discharge the second presumption, as the list of factors is so comprehensive and all-embracing.

    For illustrative purposes, consider that the list of items includes "the participation of any tax indifferent party in that arrangement". A "tax indifferent party" would, for example, include a municipality. Therefore, whenever a taxpayer concludes an agreement with a municipality, which involves a "tax benefit", then both presumptions will automatically apply despite the fact that the transaction is entered into in a normal manner and on an arm’s-length basis. How is a taxpayer justifiably expected to "defend" such a position?

    The burden of proof will shift to the taxpayer in every case, merely upon the establishment of "fact" by SARS. As such, the burden of proof created by these presumptions places all taxpayers in an unfair position, as SARS will no longer have to prove the criteria contained in the new proposed section 103.

    The presumptions of abnormality contained in the proposed section 103(2)(d) to (k), will include many aspects of most conventional financing transactions that cannot be intended to be presumed to be "abnormal". That these characteristics are deemed to indicate 'abnormality' (unless otherwise proved) can have absurd results.

    1. Legislative Interpretation
    2. Despite the concerted effort that has been demonstrated by SARS in preparing the discussion paper, which precedes the actual wording of the new proposed section 103, taxpayers may only have regard to the wording of the legislation itself in interpreting the section. Therefore, whilst being insightful and explanatory in nature, the discussion paper is of no use when uncertainties exist in the text of the proposed legislation, which will require interpretation.

      In this regard, the proposed section contains the term, "one of its main purposes", which is not defined. The resultant uncertainty that taxpayers face in interpreting what is meant by this term is disturbing, since it is expected that legislation should contain certain, precise and predictable parameters. In addition, in determining "one of its main purposes", the test that has been laid down is an objective one, rather than being subjective. It appears that a court would have to examine the facts and circumstances in determining whether tax avoidance was the sole or one of the main purposes, without any regard to the taxpayer’s ipse dixit. When this purpose is presumed, it will be impossible for a taxpayer to prove what the objective purpose actually was. The effect of including objectivity into the factual enquiry takes away the judiciary’s ability to decide on a matter based on all factors, including a taxpayer’s intention, which should take into account his ipse dixit. The decision will effectively be made by the official alleging the misconduct, as a presumption will then arise, which the taxpayer will probably not be able to rebut on facts alone, and the court will not be able to consider any evidence given by the taxpayer himself. In fact, it could be argued that the courts appear to have been stripped of their power of interpretation altogether, with their function having been effectively converted into the mere administration of tax disputes. Returning to the proposals relating to the purpose test, it is inevitable that every business decision will take into account the tax consequence of proposed business transactions, as tax constitutes a potential 29% cost. Consequently, based on the draft proposals, it may well be argued that the obtaining of a "tax benefit" is one of the purposes of every conceivable transaction. Consequently, the proposed wide purpose test, which includes "one of its main purposes" instead of "main purpose", (the latter being accepted as indicating that a dominant purpose is required), creates significant uncertainty as to whether a transaction will fail the test. There is a significant chance that this type of uncertainty may stifle a wide range of completely legitimate transactions, as taxpayers will not be in a position to determine what is permissible and what is not. Such uncertainty will have inevitable negative consequences on economic growth.


    3. Economic vs. legal substance
    4. The reference to "economic substance" as opposed to "legal substance" in the proposed section 103(2)(a) is highly problematic. The latter concept is what is usually examined when the courts examine "substance versus form" and/or "simulation", while the former concept strongly influences accounting disclosure. There are many examples where the accounting treatment (thus the "economic substance") of transactions differs dramatically from the accepted legal substance and accepted tax treatment. Examples include share-based payments, financial leases and intra-group transactions. It would create enormous confusion if the effect of the anti-avoidance proposals was that accounting treatment takes precedence over all other forms, without this being an explicit part of the Income Tax Act. We therefore suggest that either the Income Tax Act should explicitly embrace accounting treatment in all respects (and thus "economic substance") or the proposed section 103(2)(a) should instead refer to "legal substance".

      We specifically support points 2.6 and 2.7 in the BUSA/SACOB submission.

    5. The absence of an Advanced Tax Ruling system
    6. The Income Tax Act has recently been amended to provide for an Advance Tax Ruling (ATR) system. The objectives of ATR are to provide clarity in transactions and arrangements so that the differences in interpretation can be minimised. The mechanisms for bringing the ATR system into operation are still in the development stage and currently allow the Commissioner to refuse an application which is inherently factual in nature or which concerns a general or specific anti-avoidance measure. This right of refusal is not acceptable, as it denies taxpayers the right to obtain certainty regarding the application of the proposed section 103 before implementing transactions. It is strongly recommended that a linkage be provided between section 103 and the ATR system and that any contemplated amendments to section 103 be deferred until the ATR system is fully operational. In addition, for an advance ruling system to operate effectively in support of section 103, it should be compulsory that there always be a ruling given within a reasonable period of the formal ruling request.

      We specifically support point 2.3 in the BUSA/SACOB submission.

    7. A comparison to the current section 103
    8. If the purpose of a GAAR is to prevent the conclusion of transactions that incorporate "impermissible" tax avoidance, then it should be noted that the provisions of the existing section 103 are far more successful in discouraging transactions of this kind than SARS acknowledges.

      We specifically support point 1.3 in the BUSA/SACOB submission.

      The existing "sole or main purpose" test is in line with international best practice, even for fully first world economies. The introduction of a test which is based upon "one of the main purposes" is disturbing, as it creates an anti-avoidance standard for South Africa, a developing nation, which is significantly more onerous than the standard adopted by most first world countries (despite creating uncertainties, as noted in the paragraph above). As we have already noted, uncertainties of this nature have an unduly negative effect on local and foreign investment, something that South Africa simply cannot afford.

      The Commissioner is no longer required to be "satisfied" that the elements of the provision exist. In fact, there is no scope in the new provision for any discretion on the part of the Commissioner at all. When anti-avoidance is being addressed, one would hope that more than just the factual circumstances are to be examined and that the Commissioner would be empowered to consider all relevant factors before proceeding. It appears as though the section applies automatically upon the "trigger" of any of a number of random facts.

      We specifically support point 2.5 in the BUSA/SACOB submission.

    9. Retrospective application of the legislation
    10. We specifically support point 2.8 of the BUSA/SACOB submission.

    11. Deemed Abnormality
    12. Returning to section 103(2) (d)-(k), we suggest that these provisions, which deem there to be abnormality (unless otherwise proved), can have an absurd effect.

      Consider, for example, a tax-exempt entity that simply places money on deposit with a bank, which uses this as part of its normal deposit base, to lend at an interest margin. There is a "tax benefit" as the bank will get a tax deduction for the interest paid, and the recipient will not be taxed. To stress the absurdity, even if the recipient was taxed, there is still a so-called "tax benefit" to the bank, as the bank should obtain a tax deduction under section 11(a). Clearly, this tax deduction has a significant impact on the margin the bank earns (as it will be taxed on its interest income, and non-deduction of the interest paid would usually make bank lending totally uneconomic), and it can thus be said that ensuring such a tax deduction is one of the main "purposes". The flow of assets is definitely circular, as the bank must repay the exact sum of the deposit on maturity of the loan. The depositor in this example is a ‘tax indifferent party’ (but even if it is a taxable entity, subsections (d) and (k) will apply). The value of the tax deduction to the bank may well be more than the gross margin it can earn on this deposit. In summary, the existence of any one of items in subsections (d) to (k) means that the transaction is deemed to be "abnormal", and the parties must now "prove" the contrary. This can only be done in a court of law. Until they do so prove the contrary, they must treat the transaction as if section 103 applies, as the proposed new section 103 does not require the Commissioner to contend that section 103 applies, or be satisfied that it applies, it applies automatically. Thus, the bank must voluntarily disallow its legitimate interest deduction in its tax return, or risk an argument that it has material non-disclosure in its tax return, meaning that its tax return never prescribes.

      The illustration relates to a conventional lending transaction, but it could as well apply to hundreds (if not thousands) of conventional transactions entered into by businesses of all kinds on a daily basis. The end result may be absurd, but it is a very real possibility, and could undermine the whole fabric of tax compliance, with potentially disastrous results.


    13. Penalties
    14. SARS have merely alluded to penalties in their document and have indicated that penalties will be incorporated into other relevant parts of the Act. It is not fair to expect comment on one part of the legislation when the provisions relating to penalties are, as yet unknown.

    15. Conclusion

    The Banking Association is appreciative of the opportunity to add to the discussion process around the question of tax avoidance. Undoubtedly, the issues raised are complex and cannot be resolved in isolation. It does appear that section 103 would normally provide a "goal-keeping" function for those arrangements that fail to be stopped under the conventional provisions of the Act. This appearance is reinforced by the observation that most successful judgements obtained by SARS are obtained, not under section 103, but under other provisions of the Act.

    The danger of having a presumptive section 103 is that it can become an overriding provision that infers legislative intention and overrides the normal functioning of the Act. The fact that there is an anti-avoidance provision in the Act and that this provision does provide a deterrent to aggressive tax planning should not be underestimated by SARS. Furthermore, legal uncertainty is economically detrimental as it deters investment, both local and foreign. It is recommended that if there are schemes which are repugnant to the SARS, that specific legislation be introduced to prohibit those schemes. Past history bears testimony to the effectiveness of this option.


    If there are any issues which can be further explored, we would be happy to make ourselves available for discussion.


    The Banking Association South Africa.


    28 February 2006