14 March 2006

Mr N Nene

Chairman: Portfolio Committee on Finance


P O Box 15




Dear Sir




We refer to your call for comment on the discussion paper which was released for public comment via your website on 3 November 2005. It is clear that SARS has expended much time and effort in preparing the discussion paper and this is to be commended.

Set out below please find SAICA’s comments on the document.



  1. We take this opportunity to congratulate SARS on the approach taken in seeking to amend the existing General Anti-Avoidance Rules (GAAR) under section 103(1) of the Income Tax Act No. 58 of 1962 (the Act). The release of a discussion paper open for public comment on a matter of national importance, in our view, clearly indicates SARS and Government’s intention to engage all role players in the process as opposed to legislating without consultation and discussion.
  2. We fully support SARS’ intention to strengthen the existing GAAR to counter-act the various ways in which some transactions are structured or entered into which give the parties to the transaction significant tax benefits that could be questioned commercially. Equally important, in our view, is the enforcement of legislation even if this means taking matters to court, as opposed to merely passing legislation which lies dormant on our statute books.
  3. Before commenting on the proposed changes to section 103 of the Act it is appropriate to comment on a number of environmental factors that should be remedied before undertaking significant amendments to section 103 of the Act by introducing a general anti-avoidance rule as is proposed in the document. These are addressed below under "General comments".

    General comments

  5. Historically, it does not appear that SARS has relied on the existing anti-avoidance provisions containing section 103 to any great extent. It would appear that tax practitioners are more concerned about transactions falling foul of section 103 of the Act than what SARS believes the reach of section 103 to be. It is clear therefore that SARS officials appear to take a narrow view of section 103 of the Act and do not believe that it can be successfully applied on transactions that cause concern whereas tax practitioners generally take a more conservative approach and advise their clients that the transactions under consideration can be set aside because of section 103 of the Act.
  6. Furthermore, where taxpayers have concluded transactions that are not what they purport to be, SARS does not require section 103 to set such transactions aside. In such cases SARS would be correct to treat such transactions as a sham and to tax the parties as if the transactions had not been concluded at all. Where a taxpayer seeks to rely on non-disclosure and disguises the transactions to give them the appearance of something that they are not intended to be SARS does not need an anti-avoidance provision. Based on general principles of law SARS would be entitled to have the transactions struck down by the courts. It does not appear that SARS has resorted to challenging transactions on this basis to any large extent which is somewhat surprising when account is taken of the comments made by SARS and National Treasury about the types of transactions concluded by taxpayers.
  7. It is interesting to analyse the judgments of the courts in relation to section 103 and, in particular, to compare where the taxpayer succeeded with where SARS won. The vast majority of Supreme Court of Appeal decisions went in favour of the taxpayer (see the King, Geustyn, Gallagher, Hicklin, Louw (in part) and Conhage cases. On the other hand, if one looks at SARS’s successes, for example, ITC 1113, 30 SATC 8; ITC 1151, 33 SATC 133; ITC 1178, 35 SATC 29; ITC 1496, 53 SATC 229; ITC 1518, 54 SATC 113; ITC 1582, 57 SATC 27; ITC 1606, 58 SATC 328; ITC 1699, 63 SATC 175; it will be seen that the cases often involved blatant avoidance or investments in products, which deserved to be successfully attacked. One must therefore ask: "what were the essential differences between the successes and failures from SARS’ point of view?". In our view, the major lesson to be learned from the distinctions in the cases is that where the taxpayer is undertaking a genuine business or family-related transaction, undertaken in a commercial manner and properly arranged, the courts will be slow to allow section 103 to be applied even if there is a tax saving. But where the approach has little commercial purpose or is blatantly for tax-avoidance purposes, the court will support SARS in its attempts. The problem (for the fiscus), as we see it, is that where the courts have rejected the application of section 103(1) because arguably they felt that its application was not appropriate as the transactions were ordinary and bona fide, the courts may well have had to justify their approach by interpreting the relevant portions of the section so as to support their conclusion. And in doing so, they have inevitably chipped away at the edifice supporting section 103. To this end, it may well be that SARS must shoulder some responsibility, because they persisted in attacking schemes which, in the main, were commercially or family driven even though they had tax benefits; rather than confining themselves to attacking the blatant schemes. And, having procured the enactment of a new section, one wonders whether SARS will have learned their lesson, or will they attack in inappropriate circumstances (just because there has been some successful tax avoidance, as in Conhage) with the result that the courts will again seek to narrow the section’s ambit.
  8. SARS must take cognisance of the fact that taxpayers often structure their transactions because of perceived anomalies in the tax legislation, for example, interest incurred on loans utilised to purchase assets as opposed to shares/equity is allowed as a deduction. In this regard, we believe that the principles enunciated in the Drakensberg Gardens case should be codified.
  9. Taxpayers are given very limited and in some cases no tax allowances for commercial property. This should be reassessed having regard to the fact that such property is used in the carrying on of a business producing income. The grant of an allowance having regard to the commercial reality would reduce the need for taxpayers to enter into schemes such as the bare dominium schemes.
  10. Consideration also needs to be given to the introduction of "group taxation" which will also eliminate the many transactions and schemes that are developed to "artificially create" group taxation.
  11. Reportable Arrangements
    1. The rules governing reportable arrangements contained in section 76A of the Act took effect from 1 March 2005. From comments made by the Commissioner: SARS at various fora it would appear that the provisions are not having the desired effect. It is contended that if the rules governing reportable arrangements are not yielding the desired results the section should be refined and made stricter in order to yield the information that SARS seeks. It would appear that the reportable arrangements provisions in the United Kingdom and other countries are working well in informing the revenue authorities in those countries as to transactions that should be investigated. If SARS believes that the reportable arrangements provisions are not working in South Africa, they should be refined and enhanced in order to achieve the desired objective.
    2. It appears the full impact and benefit to SARS of the reportable arrangement legislation has not been taken into account in drafting the revised GAAR. In this regard, SARS needs to react swiftly by examining in detail reportable arrangements so reported and take the appropriate action rather than be inactive and seek to introduce further new legislation.

  12. Advance Tax Rulings
    1. Currently taxpayers cannot as yet call for advance rulings under Part IA of the Act as the date on which the provisions take effect must still be fixed by the President in the Government Gazette. It is important that the advance rulings procedures are activated so that taxpayers can call for rulings on various matters. We are strongly of the view that the Advance Tax Ruling System should be introduced prior to or simultaneously with the proposed new GAAR. The reason is that business requires certainty in tax as returns to shareholders cannot be reported each time contingent upon tax issues. Taxpayers will then be afforded the opportunity to obtain SARS’ view prior to the implementation of a transaction or scheme and should the taxpayer choose to pursue the transaction in the face of an adverse ruling from SARS, the taxpayer or promoter must face the full consequences.
    2. It is, however unfortunate that currently section 76G(2)(a) of the Act confers a discretion on the Commissioner to reject an application for an advance ruling where the subject matter relates to either a specific anti-avoidance provision or the general anti-avoidance provision contained in section 103 of the Act. It would be far preferable if taxpayers could approach SARS for a ruling as to whether a particular transaction falls within the provisions contained in the general anti-avoidance provision contained in section 103 of the Act.

  13. The South African fiscal legislation was enacted in 1962 and has been significantly amended and perfectively rewritten over the last number of years. It would be far preferable if the Act was consolidated and if the proposal to rewrite the Act, which was previously indicated to be on the agenda, is resuscitated in order to clean up the provisions of the Act and to harmonize the language used in the statute. It is also important that greater care is exercised in drafting the amendments to the Act. Unfortunately it is true to say that over the last few years we have seen too many amendments prepared with great haste only to be subsequently rewritten and significantly amended time and time again.
  14. The tax gap
    1. Unfortunately, the tax gap in South Africa continues to be unacceptably high when reference is made to the number of taxpayers on register in the country.
    2. According to Statistics South Africa and particularly the labour force survey covering the period September 2000 to March 2005, published on 26 September 2005, it was indicated that the number of persons actively employed in South Africa amounted to 11,9 million people at the end of March 2005.
    3. According to the 2005 SARS annual report issued by the Commissioner SARS covering the period 2003/2004 the number of individuals on register for tax purposes was 4 647 484. Unfortunately, the SARS annual report does not indicate the number of persons subject to SITE only; it is thus difficult to estimate how many people are currently in the tax system. Sometime ago it was indicated by the Commissioner that those persons paying SITE comprise some 3,5 million people. It is accepted that this is a reasonable estimate of SITE payers. This would mean that there are some 7 780 129 persons in the tax system compared to 11 000 900 persons that were economically active during March 2005. This would indicate that there are some 4 119 871 persons who are economically active in the country who are not currently paying tax in South Africa. It is important that the number of persons not registered for tax purposes is reduced as that will broaden the tax base in the country and allow government to reduce the tax rates further. It is contended that this would also result in a reduction in the levels of tax avoidance that SARS and National Treasury contends has reached unacceptable levels.


    Specific comments on the discussion paper

  15. It is enlightening to know that both SARS and National Treasury hold the view that the "vast majority of South Africans are honest, hard working and willing to pay their fair share of tax." Unfortunately some SARS officials do not always portray this view in their interactions with our members.
  16. It must be recognised that tax is a cost of doing business or earning income and as such it is legitimate to expect business to minimise its tax cost as they do with all other costs, but this should be done within the bounds of the law.
  17. It must also be accepted that in our tax system we do not have symmetry in that the tax treatment of a receipt in the hands of a recipient is not dependent on the tax treatment of the expenditure or outlay by the other taxpayer. For example, the expenditure for the purchase of a residential building is not deductible, being expenditure of a capital nature, whilst the receipt by the developer of the proceeds is taxable as income. Whilst we acknowledge that some transactions are engineered to take advantage of the non symmetry, we believe it incorrect to generalise (see last sentence under paragraph 2.2.3 Tax Planning, page 5).
  18. Whilst we acknowledge that technology has played a major role in the transmission of information across countries and continents almost instantaneously, this technology is also available to SARS to share information with other tax authorities across the globe. The Double Tax Treaties also allow for the exchange of information between countries and SARS should seek to invoke these provisions rather than seek to criticise the sharing of information by professional services firms (see paragraph 3.3).
  19. The comments on the so-called "bare dominium" schemes in part 6.2.3 on page 21 may be correct, but there are two issues that need to be raised in this regard.
    1. First, these so-called schemes have been in the market for more than 10 years, but to date very few, if any, of these schemes have been attacked in court by SARS. It is not clear as to whether there has been a reluctance on the part of SARS to challenge these schemes for fear of losing in a court of law and the consequential implications of the scheme being seen as "above board" and their subsequent increasing proliferation.
    2. Secondly, the discounting of promissory notes, which could be done to avoid receiving the rental in full upfront, is a means of raising finance on the face of promissory notes. For the rental to be received upfront and therefore being taxed upfront, the lessee will have to make payment of the full amount upfront. This does not happen factually or commercially in a bare dominium scheme.

  20. The discussion paper raises the concern in part 6.2.4 about the inconsistent treatment for tax and financial accounting. SARS needs to understand and accept that tax principles and accounting principles will never be aligned unless either financial statements are prepared in terms of the Act or the Act is amended to fall in line with accounting principles. This has not occurred in any major economy and we believe is unlikely to happen in South Africa. SARS cannot choose which accounting principles it favours where they clearly benefit SARS, but must determine tax liability solely in terms of the legislation. For example, SARS does not allow a taxpayer a deduction for a provision which is required for accounting purposes if such provision does not meet the requirements of the Act. It therefore cannot seek to tax an amount as income if such amount has been accrued for accounting purposes if such amount does not fall within the definition of "gross income". We are not aware of any of our corporate clients treating their tax department as a "profit-centre". On the contrary, the tax departments within corporate taxpayers are there to ensure compliance with the legislation. In this capacity they play a major role in improving compliance with the legislation and constantly challenge interpretations by other departments within such corporations.
  21. The notes under part 7.6 on page 37 state that "the courts in the United States have developed a variety of robust judicial doctrines to counter abusive avoidance schemes." This is an important aspect that SARS must not lose sight of in that interpretation of tax law is a key aspect in which our courts can play a crucial role. We are of the view that SARS should be more willing to take matters to court so that we too can have our courts develop a variety of judicial doctrines in this regard. SARS should explore ways to expedite matters to the courts to achieve this rather than choose to settle through the Alternate Dispute Resolution (ADR) process tax schemes which have not been challenged in court.
  22. A highly contentious issue relates to the proposed penalties for the promoters and the participants. This is a highly unique and, it must be said, regrettable approach because it effectively makes "criminals" out of bona fide operators who are able to interpret the law in a manner which SARS does not find palatable. But if the concept of penalties is to be included, a number of safeguards must be legislated for.
    1. The imposition of the penalty must be the responsibility of the Commissioner personally, and no other person must have the authority to impose this penalty.
    2. Nevertheless, there may be some grounds for the imposition of penalties on promoters of tax impermissible and evasion schemes, but caution needs to be exercised in determining who will be classified as promoters for this purpose. The term "promoter" will have to be very clearly defined so as not to result in the legislation being able to be applied to a wider population. For example, when does an accountant or a lawyer in practice, consulting his client, and who proposes a course of action to solve a particular problem, become a "promoter"? There could be a very thin dividing line between the two. If the advisor is not a promoter in the circumstances, but then using his or her knowledge or experience gained in the first assignment approaches another client in similar circumstances and advises that he or she has a solution for that second client’s problems, does the advisor become a promoter then? The term must be defined to exclude persons acting in their professional capacities by giving advice on tax for which they are paid a fee having regard solely to their time and skill as opposed to the success or otherwise of the scheme upon which they opine. Any person who receives a success fee should, in our view, fall within the definition of "promoter". Our concern is that it must be ensured that one does not pass legislation which may make professionals, for example, attorneys, advocates and tax advisors, reluctant to issue opinions on tax for fear of being fined which, in our view, will threaten the growing level of compliance in South Africa.
    3. Another contentious issue is the question of penalties for the taxpayer. It is noted that the intention is to impose the penalty "for taxpayers that substantially underreport their income" (paragraph 10 at page 48). Usually if there is an underreporting but the taxpayer has made full disclosure, penalties are not imposed. Moreover, the essence of section 76 of the Act is that additional tax is imposed where there is an intent to evade. Section 103, as a GAAR, applies only to tax avoidance, albeit impermissible avoidance, and not tax evasion. Generally section 103 is utilised when the Commissioner would be unable to succeed under the ordinary rules in the Act, for example, section 11(a), section 23(g), section 24J, and so on. The Commissioner resorts to section 103 when, but for that section, the taxpayer would be successful under the ordinary provisions of the Act. This being the case, and on the assumption that there is adequate disclosure, how can a taxpayer be put on the same level as a tax evader who, according to the OECD definition cited in paragraph 2.2.1 at pages 2 and 3, is one who entered into "illegal arrangements" and who "pays less tax than he is legally obligated to pay by hiding income or information from the tax authorities"? It is contended that where the legislation is deficient and a taxpayer has arranged his/her affairs such that they fall outside of the specific provisions of the legislation he/she should not be subject to the imposition of penalties or additional tax. As it is, SARS is already entitled to levy interest on the under payment of provisional tax and SARS cannot reverse or waive such interest under the provisions contained in section 103(6) of the Act. It is contended that the imposition of interest in such a case already constitutes a penalty and to seek a further penalty would be extreme and unwarranted.
    4. We would recommend, as a general comment, that all penalties and interest in the tax legislation should be made subject to objection and appeal, in view of the fact that penalties and interest are often imposed as a result of an interpretation of law.

  23. Application in the alternative (paragraph 10.5)
    1. It is submitted that the Commissioner should not be entitled to raise section 103 as an alternative basis for assessment as that would be contrary to the taxpayer’s right to administrative justice contained in section 33 of the Constitution and as reinforced under the provisions of the Promotion of Administrative Justice Act No. 3 of 2000. The Commissioner: SARS should be required to assess a taxpayer utilising the general provisions of the Act and if he is unable to do so, SARS should then invoke the general anti-avoidance provision. However, it is contended that to allow the Commissioner to rely on specific provisions in the Act and to rely on the general anti-avoidance rule as a catch all provision would undermine the integrity of the tax system in the country.
    2. If the Commissioner were to be entitled to rely on section 103 of the Act as an alternative basis for assessment it would clearly indicate that the Commissioner is not satisfied that the section should apply, in that he is seeking to rely on the general provisions of the Act.
    3. In dismissing the application of section 103 in the alternative the discussion paper says as follows:
    4. "At the same time, the arrangement may be so artificial and contrived that the application of section 103 may also be appropriate (in the event that the arrangement does avoid section 24J on technical grounds)."

    5. Where the arrangement is so artificial and contrived the Commissioner should not be required to rely on section 103 but should be entitled to rely on general principles of law in setting aside the transaction on the basis that the transactions is a sham.
    6. If the application in the alternative principle is adopted we caution that it must be ensured in its application that it is not automatically relied upon by SARS as a "catch-all" section of last resort without due and proper consideration. It is for this reason that we suggest that transactions or schemes that may be challenged by SARS in terms of section 103(1) of the Act, be assessed by a specialist department within SARS as opposed to individual assessors making the final decision. These specialist departments could be regionally located with a central specialist department to consider all cases in terms of section 103(1) going on appeal.


    Proposed new section 103

  24. The proposed section does not adhere to the concepts of certainty, simplicity and neutrality, which have been cited as fundamental requirements for the effectiveness of tax law.
  25. Whilst we understand and share the concerns by SARS that the existing "sole or main purpose" test falls short in successfully attacking tax schemes, we are of the view that the proposed "sole or one of its main purposes.." test is too far reaching and has the real risk of bringing the vast majority of unintended transactions within the ambit of the proposed GAAR.
  26. By way of example, assume a taxpayer can earn 5% interest on his or her savings and also has the opportunity of earning the equivalent or slightly lower percentage return by investing in listed preference shares. If the taxpayer chooses to invest in preference shares, there can be no doubt that one of the main reasons for such investment is the fact that dividends are tax free. Whilst we acknowledge that the simple fact that one of the main purposes is the obtaining of a tax benefit, the other requirements of normality will have to be satisfied, the unnecessary onus placed on the taxpayer to satisfy an assessor in this regard is in our view unnecessary.
  27. A further example is where the taxpayer could choose to purchase fixed assets outright for cash paid in full upfront or obtain the use of the asset in terms of a lease. The purchase of the asset will give the taxpayer a wear and tear allowance over a period of, say, 10 years whilst the lease of the asset over, say, a 3 year period will give the taxpayer a write-off of the lease payments over the period during which they are paid. Ignoring the provisions of recoupment at the termination of the lease, the taxpayer clearly has a distinct advantage to obtain by deferring tax through leasing the asset over a shorter period than that over which a wear and tear allowance will be granted.
  28. While we are mindful of the fact that this is how section 103(1) read prior to 1978, the expression "sole or one of the main purposes" is grammatically impossible. The Shorter Oxford English Dictionary defines "main" as meaning "chief in size or extent; constituting the bulk; the chief part; of pre-eminent importance; principal, chief, leading". (Admittedly there are some other more metaphorical meanings such as "great in numbers, of great size or bulk" which does not require a majority, but the term "main" is universally interpreted in the Act to mean more than 50%, such as where a manufacturing building must be "wholly or mainly used … for the purpose of carrying on therein any process of manufacture".) Webster’s Unabridged Dictionary gives similar meanings and emphasises the more than 50% type of approach.
  29. It is believed that having both the sole or main purpose widens the ambit of the section to the distinct detriment of taxpayer rights. It appears as though SARS wants to apply section 103 to a secondary purpose and all commercial transactions would have as a secondary purpose an element of tax efficiency. We submit that the objective purpose requirement should include reference to arrangements or transactions which are contrived, artificial or have an element of pretence.
  30. The same problem appears in the Australian legislation where the test is the sole or dominant purpose – "dominant" has much the same meaning as "main", but one does not know if the word "dominant" is used elsewhere in their legislation to mean more than 50%. The New Zealand legislation deals with it in the negative by applying the GAAR only where the tax benefit was something other than an incidental benefit, and maybe that makes more sense (though it is a much more difficult standard to meet). Maybe the test should rather be that the tax benefit is the sole or main purpose or, if not the sole or main purpose, is a purpose of such relevance or weight or importance that it has a relevance or weight or importance equal or substantially equal to at least one other purpose which does not involve the obtaining of a tax benefit.
  31. We submit that paragraph (a) of section 103(1) should be amended as follows:
  32. Has as its sole or one of its main purposes, the obtaining of a tax benefit "other than that expressly provided for in the Income Tax Act".

  33. This amendment would serve to avoid the application of section 103 in instances where the taxpayer has legitimately utilised sections of the Act which have the purpose of deferring taxation liabilities.
  34. The proposed section 103 allows SARS to adjust any step or part of a scheme where that step or part is deemed to fall within the legislation. When SARS makes a section 103 adjustment to a part or a step of a transaction, will SARS then look at the entire scheme and all parties to a scheme to ensure that no double taxation results from the application of section 103? It is not clear whether a SARS adjustment or re-determination to part or a step in an arrangement will result in another section of the arrangement being re-characterised as abnormal.
  35. In many instances it may be impossible to have knowledge of all the steps to all transactions that may occur from a simple investment. Who then will be liable for the tax as a result of any adjustment made by SARS?
  36. We urge the Commissioner to consider introducing a discretionary element to the abnormality requirement and the purpose requirement. This would entail a so-called "The Commissioner must be satisfied" requirement so that taxpayers are not immediately characterised as falling within the section i.e. the Commissioner would be required to apply his mind as to whether these elements are present as part of an avoidance scheme. This would allow for taxpayers to apply the Promotion of Administrative Justice Act No. 3 of 2000 to any decisions which are made by SARS officials which negatively impact on the rights of taxpayers.
  37. Fundamentally, although the objective purpose test is welcomed, the objective normality test is extremely far reaching, and will impact on normal business transactions which involve no contrivance or pretence. It is mentioned in the discussion paper that inter alia Australia’s objective purpose test was evaluated, and in fact some eight of the factors used in their equivalent GAAR were used in the proposed abnormality tests (almost verbatim). In Australia, the guidelines set out an indication of what might be considered to constitute an avoidance scheme with the purpose of avoiding tax. Surely, on this basis the criteria should fall within the definition of "arrangement" or alternatively, simply be set guideline in an Interpretation Note.
  38. Instead of listing features of impermissible tax transactions per the abnormality tests, we would be better served to use the "sniper approach" aimed at specific avoidance schemes. For example, the amendment to section 24J which prevented the adjusted initial amount from being inflated in circumstances where the issuer made payments directly or indirectly to the holder or connected person of the holder (Revenue Laws Amendment Act No. 32 of 2004).
  39. We strongly believe that the list of factors listed in terms of sub-section (2)(a) to (k) should not form part of the legislation but be included in an interpretation note. Our reason for the exclusion is that there is a real possibility that the list will result in taxpayers having to rebut the abnormality presumption merely because one of the factors listed are present without regard to the relevant facts and circumstances. This view is reinforced in terms of section 103(4)(b) where the existence of any one of the factors listed in paragraphs (d) to (k) will be presumed to satisfy the "means or manner" test.
  40. We are further of the view that the objective test should not be applied in isolation without regard to the purpose as evidenced by the actions of the taxpayer amongst other factors.
  41. It is noted that the tests used in Australia, as referred to in footnote 93 at page 29, have been incorporated into section 103(2). In Australia, however, the tests are used to determine objectively whether the scheme was entered into for the sole or dominant purpose of tax avoidance, whereas in the draft section 103 it is utilised for determining abnormality. Moreover, the Australian legislation looks at all of the issues whereas our draft is happy that only one is present.
  42. The proposed section 103 places the onus of disproving the abnormality requirement in the taxpayer’s hands without any discretion being exercised by the Commissioner in applying his mind to the transaction or arrangement. Any transaction or arrangement which falls within paragraphs (a) to (k) of section 103(2) appear to automatically be abnormal in terms of the guidelines contained in that section. As a result, the taxpayer is effectively immediately presumed guilty of a contravention even for transactions which would normally not be considered to be "abnormal".
  43. It is not clear whether SARS is required to apply their mind when looking at the abnormality requirements contained in the preamble to section 103(2). When one looks at the "means or manner" of the arrangement, the section refers to the fact that each step or part must be determined objectively with reference to the relevant facts or circumstances. Although this is broadly worded, it is unclear whether SARS is required to apply their mind based "objectively on facts and circumstances" of each case. It is furthermore uncertain whether SARS will apply this section as a catch-all to include all transactions that comply with the criteria set out, given the wording "including but not limited to" in the preamble of section 103(2). It is possible to interpret the latter wording as implying that SARS should exercise their discretion, where the arrangement or transaction falls outside of the guidelines referred to in sub-sections (a) to (k). As the latter sections are widely worded to capture any bona fide transactions, including the exercise of the Commissioner’s discretion in the preamble to section 103(2) would ensure that taxpayer’s rights are not eroded by SARS auditors whose aim is achieving certain revenue targets.
  44. Section 103(2)(b) refers to "the time at which the arrangement or any step therein or part thereof …" and "the length of the period during which the arrangement or step was carried out". This sub-paragraph is very vague with regard to timing and does not indicate how the Commissioner will objectively apply this test in terms of the facts and circumstances to the arrangement in question. If sub-paragraph (b) is applied strictly by the revenue authorities, ordinary transactions which are not contrived or artificial will fall within the anti-avoidance provision.
  45. Sub-paragraph (d) of section 103(2) refers to "any circular flow of cash or assets between or among parties to an arrangement". This circular flow may apply to bona fide commercial transactions which are not aimed at tax avoidance. A circular flow will arise, for example, where a taxpayer chooses to acquire a motor vehicle through a finance lease or directly with a loan from a commercial bank. The taxpayer borrows money from the bank or finance house and repays the loan with a possible different interest rate which is below the prime rate of interest, usually charged by commercial banks. A similar circular flow may arise where an asset is disposed of to a family trust. This circular flow of cash or assets among parties is part and parcel of everyday transactions which, based on the wording of the section, are now determined to be abnormal. It is submitted that legitimate tax planning transactions or arrangements are now brought within the ambit of the anti-avoidance section by virtue of paragraph (d) of section 103(2).
  46. Section 103(2)(e) refers to the participation of any tax indifferent party to that arrangement. "Tax indifferent party" is defined as:
  47. "(a) any person that is not subject to any tax imposed by this act;

    (b) any person that participated in an arrangement in such a way that the amount derived by that person in connection with the arrangement is substantially matched or offset by any expenditure or loss incurred in connection with the arrangement; or

    (c) any company or other entity which is established specifically for the purpose of participating in that arrangement."

    Paragraph (a) is very broad in that it would include any transaction with a non-resident that has no South African source income.

    Paragraph (b) of the above definition is very widely drafted and will include any taxpayer that receives income and has incurred expenses in earning that income. If paragraph (b) is applied strictly by SARS, ordinary transactions which are not contrived or artificial will necessarily fall within the anti-avoidance provision.

    The application of paragraph (c) of the above definition may result in the application of the anti-avoidance provisions where a company is formed to acquire another business. The latter transaction is certainly not aimed at tax evasion but rather at limited risks and liabilities for the shareholders. Legitimate business transactions are therefore brought within the ambit of the anti avoidance legislation, without taxpayers having any certainty or clarity as to how SARS will apply these guidelines.

    The taxpayer has no comfort as to the manner in which SARS will apply section 103 or whether SARS would "apply its mind" reasonably with regard to the application of the abnormality guidelines.

  48. Sub-paragraphs (f) and (i) of section 103 (2) appear to contain similar implications and could perhaps be combined or alternatively included in part (b) of the "tax indifferent party" definition. The only difference that can be ascertained between paragraphs (f) and (i) is the difference in the wording of financial position (purely monetary amounts) and economic position (refers to a broader concept than financial amounts).
  49. Sub-paragraph (g) of section 103 (2) refers to the inconsistent treatment of amounts for tax purposes by parties to the arrangement. There is no requirement for consistency of treatment for different taxpayers in the Act or in legal precedent. A motor dealer for instance would treat a motor car as his stock-in-trade whilst another taxpayer may treat a vehicle purchased as a fixed asset. Both these bona fide business transactions are now brought within the ambit of the anti-avoidance legislation through paragraph (g). The present legislation and legal precedent recognizes the need for a different tax treatment by different parties to a transaction or arrangement. We submit that section 103(2)(g) should include reference to transactions where there has been some form of pretence, artificiality or contrivance in the creation of a transaction.
  50. Sub-paragraph (h) of section 103(2) refers to parties which do not deal at arms length whether they are connected or not. The Act and legislation generally implies that all independent parties will act at arms length as there is no rationale for the inflation or reduction of the price beyond market value between independent parties to a transaction. If the purpose or intent of the section is to avoid collusion between independent parties which transact above market value, then the section should refer to collusion between independent parties. SARS will, however, collect their additional taxes from the party to the transaction that makes a windfall where the counterparty has overpaid in relation to the "market price". This section currently captures most bona fide commercial transactions. We submit that in order to exclude commercial transactions, the proposed section should include reference to the fact that SARS must be satisfied that where the transactions are concluded for more than market value for any assets or services involved in that arrangement, the parties are not independent and or collusion has taken place between those parties.
  51. Neither paragraph (j) nor (k) is included in the Australian GAAR legislation. It is not clear as to how SARS will interpret or apply paragraph (j) which provides for a reasonable expectation of a pre-tax profit in connection with an arrangement. Must taxpayers draft a business plan for transactions which they have entered into, and will transactions which are loss leaders aimed at gaining market share or attracting customers to acquire other more profitable products, fall within the ambit of these provisions? It is also not certain when the calculation for the pre-tax profit will be determined, and whether this is overall pre-taxed profit or based on the yearly performance of the transaction or arrangement. An estimated pre-tax profit may not crystallise due to unforeseen events during the course of a transaction which may increase the costs incurred in terms of the transaction. This sub-paragraph does not appear to recognise the impact which such events may have on the pre-taxed profit of a transaction, arrangement or scheme.
  52. Sub-paragraph (k) is also unclear in relation to how the pre-taxed profit must be determined in relation to the tax benefit resulting from the arrangement and appears to override existing legislation. The fact that the legislation is harsher than other countries, may act as a disincentive to foreign invest in South Africa.
  53. We are uncertain as to how regional SARS offices will apply section 103 given that the onus of disproving both the abnormality and purpose requirements is now placed on the taxpayer. Where any additional section 103 disclosures are required to be provided by the taxpayer in the income tax returns, the taxpayer’s rights may be significantly and unfairly eroded should the information provided fall within the abnormality guidelines.
  54. On a technical and interpretational level, there is a difficulty that we would like to highlight:
  55. The proposed section 103(3) requires that the purpose must be determined objectively by reference to the relevant facts and circumstances. It is interesting to note that in SIR v Gallagher, 40 SATC 39, the Supreme Court of Appeal explicitly rejected this approach when it was proposed by the Commissioner’s representative. The reasoning of the court was that one had to use a subjective test, i.e. determine what the taxpayer had in mind, in order to determine the purpose. If one uses an objective test, according to the court, there would be no real distinction between effect and purpose.

  56. In terms of the proposed paragraph 103(4), whenever a tax benefit is proved by SARS, the purpose and abnormality requirements now appear to be automatically fulfilled. The onus will therefore rest on the taxpayer to disprove the purpose and abnormality. Our concern is that for example where a taxpayer has a choice between entering into a finance lease with a lower interest rate or an outright acquisition, these transactions may be classified as abnormal in terms of the present guidelines. These bona fide business transactions fall foul of the abnormality requirements as provided in the proposed section 103(2).
  57. Many transactions require the use of a Special Purpose Vehicle/Entity (SPV) but are not impermissible transactions within the intended meaning. An example of such a circumstance is the SPV in a securitisation structure where the SPV is used solely to ensure insolvency remoteness from the lender’s perspective. The proposed sub-section 4(b) now requires the taxpayer to prove that the means and manner is that which would normally be employed for bona fide business purposes. Having regard to the fact that the widespread use of this structure will no longer satisfy this requirement, we believe this enormous burden on the taxpayer will stifle business competitiveness from a global perspective. The mere use of a tax indifferent party does not automatically connote tax avoidance and whilst we do agree that many such structures use SPVs for this purpose, the generalisation should be avoided. The further concern is that funding transactions for local government will automatically be affected merely because local government is not liable for income tax.
  58. We are concerned that the abnormality guidelines are too broad in their determination and may lead to absurd results. We would urge SARS to utilise the current mechanism of reportable arrangement and advanced rulings as preferable mechanisms to early identification of transactions or arrangements which would fall foul of the anti-avoidance legislation. These sections are currently applicable and will identify any abusive transactions, although the current turnaround times for section 76A scheme reference numbers are time consuming. The key, however, is that SARS need to study reported transactions timeously so that they can act to prevent the widespread use of the so-called impermissible transactions.
  59. The exclusion of having regard to the parties being connected persons when determining whether they are dealing at arm’s length has far-reaching and, in our view, unintended consequences. For example, interest free loans between group companies could now be successfully attacked on the basis that the non-charging of interest results in the avoidance of tax and it is deemed to be carried out by means or in a manner not normally employed for bona fide business purpose in terms of the proposed sub-section 103(4)(b). We are firmly of the view that the relationship of the parties should be totally ignored.
  60. The onus to prove the normality of the various transactions should not, in our view, fall upon the taxpayer but the Commissioner should be required to prove abnormality if the Commissioner is invoking section 103.
  61. Under the proposed sub-section (6) the Commissioner may apply the anti-avoidance rule in the alternative for or in addition to any other basis for raising the assessment and it is contended that this is unacceptable for the reasons given above. As discussed above, the Commissioner should be obliged to rely on the general principles of the Act in subjecting a taxpayer to tax and should not be permitted to use section 103 as means of extending the reach of the fiscal legislation, as that power should only reside in Parliament itself.
  62. Regarding the date of implementation per point three of the notes to Annexure F of the discussion paper, it appears via the wording that it is possible for the legislation to be retro-active. Although we were advised at the breakfast launch of the legislation that the amendments would not be retroactive, we strongly urge SARS not to enact the legislation retrospectively as this word significantly erodes taxpayer rights.

    The Constitution

  64. The discussion paper does not appear to consider the constitutional issues relating to a general anti-avoidance rule. GS Cooper’s "Tax Avoidance and the Rule of Law" comments on the constitutional validity of a general anti-avoidance rule. In chapter 9 of Cooper’s work L Mutén states the following at page 307 thereof:
  65. "Meanwhile, in a doctoral dissertation presented in 1995 by Anders Hultqvist, the issue has been raised whether the GAAR is constitutional. The argument against is that the GAAR in its present form authorises interpretation of the tax law by knowledge and by analogy, something that might violate the constitutional provision, under which taxation can be imposed by law only."

  66. A concern that arises is that the general anti-avoidance rule that is proposed to be introduced shifts the power to tax from Parliament to SARS.
  67. In a constitutional democracy, the number of discretions conferred upon the Commissioner: SARS should be reduced and restricted wherever possible. By introducing a GAAR, the danger is that the authority of Parliament will be undermined and greater taxing power conferred on SARS such that SARS can seek to tax transactions in any manner it sees fit under the provisions of the GAAR.
  68. It is recommended that the whole system of imposition of additional tax and/or penalties in South Africa should be reviewed and refined. In this regard reference must be made to the comments found in Duncan Bentley’s "Taxpayers Rights’ * An International Perspective" (the Revenue Law Journal, School of Law Bond University, Queensland, Australia, 1998) where reference is made to the system of penalties levied by the Inland Revenue Department of New Zealand. In that country the degree of penalties depends upon the alleged offence committed by a taxpayer and South Africa should be reviewing the penalty provisions and amending them such that taxpayers know what level of penalty can be expected depending upon the nature of offence committed, for example where a taxpayer inadvertently omits income from their tax return or commits a fraud on SARS.


  70. Whilst we acknowledge the need to amend the current GAAR, our overriding concern is that, despite the discussion paper’s stated intention to the contrary, the proposed new legislation is so wide in its scope and so far-reaching that it is going to affect ordinary, everyday business transactions, i.e. it will introduce uncertainty even into, what have hitherto been, "vanilla" transactions, because there is the possibility that section 103 will apply.
  71. To illustrate the point, take the example of a purchaser wishing to buy an existing business operated by a private company, with the purchaser having to utilise bank funding to finance a portion of the purchase price.
  72. The purchaser has a choice of buying either the shares in the company or the business out of the company and in the latter case will often form a new company to buy the business. There are usually three reasons for buying the business in these circumstances, namely:

It is clear that two of the three reasons are tax-motivated and therefore tax avoidance would be, if not the sole purpose, certainly one of the main purposes of structuring the transaction as a purchase of a business.

  1. Thus one has an "arrangement" which has the effect of obtaining a "tax benefit" and the sole, or at least one of the main purposes, was the obtaining of a tax benefit, objectively determined. Consequently, there is a rebuttable presumption that the final requirement of abnormality exists. Hence, section 103 is prima facie applicable.
  2. This example demonstrates how easy it will be for section 103 to bring about widespread uncertainty in regard to so many ordinary and everyday transactions. And one can quote many other examples. One of them might be the situation where a company agrees to pay a cash bonus to its employee on the assumption that the employee will use the after-tax amount to subscribe for shares in the employer company. Here one has a circular flow of cash so that, again, subject to the ability to rebut, all of the requirements for section 103 have been met (this is required by the fact that SARS still does not accept that a company can actually incur expenses in the form of a bonus when it allots its own shares to the employee). This example in part also demonstrates a problem which taxpayers generally have to deal with, and that is that the Income Tax Act all too often has not kept pace with the change in the way commercial transactions are structured and accounted for. If share-based payments are a normal and everyday feature of the commercial world, why does the Act not allow a deduction?
  3. While the net of the proposed section 103 has certainly been cast wide enough to capture certain kinds of schemes, it has been drafted so widely, as indicated, that it will capture many more where there is economic substance and genuine outflow of funds.
  4. We would therefore like to suggest the following:
    1. The comment referred to in paragraph 2.2.2 (page 3 – 4) that the objective is to deal with impermissible tax avoidance, being artificial or contrived arrangements with little or no actual economic impact upon the taxpayer, etc, should be written in to the section as a guide for determining when the section ought to be applied.
    2. Similarly, if most abusive schemes typically involve little or no economic risk, there should be a safe harbour so that if there is economic substance and/or economic risk, the section cannot apply.
    3. Given the large number of circumstances which give rise to a rebuttable presumption of abnormality, is it not too wide-ranging to have the rebuttable presumption arise when only one of the criteria is in place? Should the minimum threshold not be that at least any two or three of them apply? Even better, while the list of criteria can be included as factors to be taken into account, the inclusion of a rebuttable presumption is a step too far. (The inclusion of the list of indicia for abnormality, coupled with the rebuttable presumption, is, as we understand it, intended to reduce the burden on SARS of having to tackle each individual scheme separately on a case-by-case basis, even though the schemes’ essential elements are identical (see paragraph 10.2 at page 50). But we are not convinced that this will have that effect. Currently taxpayers, as part of their objection, will submit that the abnormality does not exist. Now, even though the section requires that there is abnormality to be presumed until the contrary is proved, the objection will simply submit that the contrary has been proved. So how have circumstances changed?)
    4. If the above recommendations are not written into the law (though we would urge that this be the case), at the very least they should be contained in an interpretation note issued simultaneously with the legislation so as to show how SARS will apply the section. We stress "simultaneously" because all too often new sections are enacted and it can take several years before the interpretation note is written, leaving taxpayers with significant uncertainty in the interim.

  5. The power which the proposed section 103 carries and, hence, its possible adverse effect on ordinary transactions is encapsulated in the comment made in paragraph 10.3 at page 56, where the following is stated:
  6. "It is anticipated - and intended - that it will be very difficult for a taxpayer to rebut the presumption of a "tax avoidance" purpose in any case in which a tax effect has been established and the taxpayer has been unable to rebut a presumption of "abnormality" arising under the proposed amendment to the Abnormality Requirement." (our emphasis.)

  7. We urge SARS to consider introducing a requirement relating to the abnormality and purpose sections, which require the Commissioner to be "satisfied" that an abnormality has arisen or the purpose is to avoid tax. Although we are supportive of GAAR legislation, we consider that the broad wording of the section results in too many bona fide commercial transactions falling within the ambit of anti-avoidance legislation. We are aware that the Australian Tax Office has recently introduced new interpretations for their GAAR legislation which reduces the application of the legislation to bona fide commercial transactions. In the light of such relaxation, we consider that making South African GAAR legislation harsher than in other countries may be imprudent and injudicious.
  8. Concern has been expressed regarding the cost taxpayers will incur in defending section 103 claims by SARS auditors. It has been suggested that SARS consider providing assistance in bearing some or all the additional cost burden incurred by the taxpayer in defending a section 103 attack in order to ensure the application of the section by SARS only when warranted and justified. The purpose of this requirement is to ensure that regional offices do not abuse the proposed section 103, in order to achieve their revenue targets. Alternatively, section 103 cases should be submitted to an independent panel for scrutiny where both the SARS auditor and the taxpayer may present their cases as happens, for example, in Australia. Similar types of independent panels currently exist in other international jurisdictions and appear to successfully reduce the costs of disputed transactions. Such independent panel persons may also be also be usefully applied to improve the limited success of the Alternate Dispute Resolution (ADR) process.
  9. If the proposed GAAR is introduced SARS should be obliged to ensure that it does not recover any greater amount of tax by applying the rule than it would otherwise have recovered. The concern raised here is that SARS could seek to re-categorise dividends in the hands of one taxpayer as interest thereby increasing the tax payable by one taxpayer without granting a deduction to the entity paying the dividend. It is unfortunate that in certain specific anti-avoidance provisions we already have such an effect. Thus under section 8E of the Act, for example, SARS can tax certain dividends as interest in the hands of the recipient only but such dividends will retain their nature as dividends in the hands of the entity paying the dividends. In invoking the general anti-avoidance rule SARS should be required to take account of all parties involved in the transaction and where SARS seeks to increase tax payable by one party by increasing the income attributable to that party it should be required to take account of the fiscal consequences facing other parties to the transaction.
  10. We stress the need to ensure that in the quest to counteract so called impermissible tax avoidance transactions/schemes, the legislation does not also counteract permissible tax planning in mitigation of tax which will adversely affect business in South Africa, especially when we find ourselves in a higher interest rate environment in the future.
  11. Finally, regarding the date of implementation per point 3 of the notes to Annexure F of the discussion paper, it appears via the wording that it is possible for the legislation to be retro-active. This is unacceptable and unconstitutional and is objected to in the strongest possible terms. The effective date should only be applied prospective to new transactions entered on or after a specified date. The proposed wording of "or another specified date" makes this unclear and should be changed. Furthermore, if any step in an existing transaction is altered such a change should not be subject to the new legislation e.g. entering into an interest rate swap. We would also recommend that, in future, wherever possible, all tax amendments should be made effective from the beginning of a year of assessment. This would avoid much of the confusion that is caused when effective dates are interspersed throughout each year.
  12. We appreciate the opportunity to provide input on the proposed anti-avoidance legislation. Our members in principle support both the intention and principles of GAAR legislation which is designed to deal with a small percentage of unscrupulous taxpayers and promoters, but due care must be given to ensure that the legalisation achieves this objective effectively. We are of the view that compliance by taxpayers with the help of tax practitioners has seen a dramatic improvement in the last few years.


Please do not hesitate to contact me should you require further information.

Yours faithfully


J Arendse N Nalliah


The South African Institute of Chartered Accountants