LISPA is the representative body for Linked Investment Services Providers (Lisps). A Lisp is a type of asset manager which specialises in administering client investments, of various kinds, as well as other financial products. In doing so it is regulated by the Financial Service Board in terms of the Financial Markets Control Act and the Stock Exchanges Control Act.

As at 30 September 2000 LISPA had 16 members, administering some 318 000 client accounts with a total value of some R 58,8 billion comprising mainly units in unit trusts. This holding represents almost 50% of the value of the Unit Trust Industry, which is administered by Lisps. It is therefore a significant financial services industry in the RSA.

Since the business of a Lisp is administratively intensive and complex, it stands to reason that a new tax, like Capital Gains Tax (CGT) will have a strong impact on the business carried on by a Lisp and that the introduction thereof will affect the ability of a Lisp to administer client investments subject to such a new tax. LISPA’s comments are made in four parts and should be seen against this background.

Appropriateness of CGT
LISPA is not convinced that CGT is an appropriate tax to be imposed in the RSA at this stage. The reasons for imposing CGT have never been fully explained nor has there been opportunity for a proper debate on its desirability. Whilst we accept that it remains Government’s prerogative to impose taxes, it does not follow that affected parties should not be thoroughly informed and consulted. A proper consultative process contributes to a better understanding by both sides of the reasons for and against impositioning of a new tax such as CGT. We therefore strongly urge that the need for CGT in the RSA and the need and benefit thereof be referred for wider public debate so as to achieve an outcome that all interested parties accept more readily than is currently the case.

The foregoing is given as a basic point of departure and our preferred route forward.

LISPA accepts the possibility that the foregoing point of view may not prevail and that CGT, adjusted somewhat from its current proposed format, may be implemented. We commit ourselves to co-operate to the fullest extent possible so as to ensure that the ultimate result is the best that it could possibly be, given the circumstances.

2. Principles for CGT
LISPA submits that CGT and its introduction should at least comply with the following principles, which are for present purposes separated into those relating to the process of its development and those relating to issues of substance.

2.1 Process
The process followed in developing CGT must be transparent and as inclusive as possible. LISPA is dissatisfied with the process that was followed, as will appear later.
Those who stand to be affected by the tax should have sufficient time to prepare themselves to administer and deal with the tax. LISPA members have to be ready to administer CGT from the first day that it is introduced. At present LISPA has a very real fear that the 1 April 2001 deadline for imposing CGT will not be moved and that its members will be non-compliant from that day.
When legislation is distributed for comment, it should be as comprehensive as possible, failing which the commentator is unable to appreciate the full extent of, and context within which the proposals stand that are being commented upon. The absence of the regulations to be promulgated in terms of the provisions of the Bill is thus a serious shortcoming.

2.2 Substantive Principles

The tax must only tax real gains (increases) in capital, which requires that some form of indexation be introduced to enable the calculation of the real gain. The Bill does not provide for indexation. The failure to introduce indexation has the effect that capital appreciation at a rate lower than the tax rate will result in CGT becoming a tax on capital, which is not what was intended. This will lead to capital diminution and the reduction of taxpayer wealth.

The tax should also be easily understood and every effort should be made to ease its administration so as to keep down the cost of compliance and the cost of collection. This will maximise the tax take and reduce upward pressure in future on the tax rate. Although the Bill is written in a clear and simple style, it suffers from excessive complexity in relation to a number of issues, such as the determination of the base cost of assets, (as will appear later).

SARS has an obligation to inform, explain and educate taxpayers on the taxes that it develops and implements. There has been no such initiative by SARS on any level. Even those knowledgeable about tax require an explanation on various aspects of a new tax Bill. This is usually done in the Explanatory Memorandum. Unfortunately even the Explanatory Memorandum which accompanied the Bill is very brief and does not assist in improving an understanding of the provisions of the Bill. Furthermore, the deficiencies in the consultative process make it difficult to gauge the extent to which the tax is generally understood. A better development process will increase the level of understanding. This is not the case in relation to this Bill and the process whereby it was developed.

The financial system functions with various forces and factors in equilibrium at any given time. A new tax may easily cause a new equilibrium to come into effect, as a result of which certain industries are competitively disadvantaged, and the flow of investment funds in the financial sector is altered as new and innovative products are developed to minimise the impact of the new tax. Such upheaval should be avoided if at all possible. LISPA considers the fundamental design of the Bill to be reasonably sound in this regard but cautions that a lot of the matters that will determine the impact of CGT will only emerge from the detailed provisions of the Bill. Some of which are still uncertain or unclear. LISPA raises numerous points of detail in the commentary below that are still outstanding and we are aware that other financial industries hold a similar view. It is thus necessary that the outstanding detail be settled as soon as possible.

The collection of tax has various costs associated with it. These costs fall out into two categories, being the direct costs incurred by the tax collector (SARS) and the indirect cost of compliance, incurred by the taxpayer (including the industry administering the tax). In this case, the second category of costs is unknown, expected to be significant and is mostly deductible from income. Preparing for and complying with CGT thus reduces the income tax take of SARS accordingly. Initially SARS did publish some figures in which the expected tax yield was disclosed, but neither the first nor the second category costs were discussed at all. It should be a principle of any new tax law that a verifiable, comprehensive cost/benefit analysis should be made and disclosed to enable an objective assessment of the proposal.

3. Time available for Comments

The Minister of Finance announced Government's intention to introduce Capital Gains Tax (CGT) during his budget speech in February 2000. This was accompanied by an explanatory brochure which explained some basic elements of the proposed tax. Comments were then invited on the proposals and were provided. Subsequently LISPA met with representatives of SARS on three occasions during 2000. On these occasions the nature of the business of LISPA members was explained. It was also explained that LISPs, as a type of asset managers, administer client investments. Both sides recognized that CGT would introduce many new complex issues into the business of LISPA members, and would affect their business and products significantly. It was understood that these issues had to be resolved and doing so would require settling many varied points of detail, which then have to be taken up in clear legislative proposals, to enable members to effect changes to their computer systems and business practices.

Until the publication of the Bill, no further official information was made available on many of these issues. Despite undertakings to publish more detail earlier, the current Bill was only published on 12 December 2000 and time for comment was a scant 30 days, stretching over the Christmas and New Year holidays, a period during which a significant number of key people were away on holiday. This severely restricted LISPA's ability to respond on an integrated basis.

Both the period allowed for comment as well as the timing of that period were inappropriate and taints the consultative process accordingly. LISPA submits that this part of the consultative process was inadequate and unsatisfactory.

Nature of Consultation
CGT is a new tax, with wide application and many consequences. Its successful implementation requires that it has to be clear and that all those affected understand it sufficiently to comply with it. In this way its introduction can be likened to the introduction of VAT. However, the process followed in developing the VAT Act stretched over a longer period, was more transparent and more inclusive. Measured against that process, the consultation in respect of the CGT proposal seems to have been concluded superficially and with indecent haste.

LISPA submits that the consultative process followed to date is inadequate in respect of its transparency and its inclusivity of those parties directly affected thereby. In addition it is qualitatively inadequate since it has failed to offer an opportunity to discuss, debate and resolve numerous matters of detail that are still outstanding. We do record that the individuals of SARS whom we consulted with were very helpful and showed understanding, but that cannot cure the fundamental flaw that the process was not comprehensive, inclusive, transparent and detailed enough. Instead of fostering acceptance, it has created uncertainty and suspicion.

LISPA submits that more discussion and debate is required on many issues, as will appear from the enclosed comments, before the proposals should be cast into law. That will not only ensure a better understanding, but also enable better compliance through the removal of outstanding issues.
5. Process to Completion and Implementation Date

SARS is committed to the date of 1 April 2001 as the due date for the implementation of CGT. That seems to be the main reason for the unseemly haste with which the proposals are being forced forward, although the significance of that date has never been explained. This means that the Bill will have to be passed by Parliament before that date. Given that, it implies that there will be slightly more than two months after the public hearings in the Portfolio Committee to consider and implement the final provisions of the Bill.

LISPA emphatically submits that it will be impossible to comply with the legislation on this date. There are two sets of reasons for that:

Firstly, there are many outstanding matters that have to be settled before LISPs will be able to know exactly how they have to change and adapt their systems and business practices. These are not settled at present and may only be settled once the final Bill is prepared, a process from which we were excluded. Not only are we now excluded from further stages of the Bill’s development, but we are not even sure that the issues we raised will be addressed or, if they are addressed, that members will be able to comply with the result.

Secondly, even if all outstanding issues are settled immediately, members will require more than the time available to 1 April 2001 to effect the changes to their computer systems and business practices. LISPs have to have fully functional systems on the day the CGT becomes effective to be able to administer their client investments and report thereon from that day. A very sincere effort was made by LISPA to persuade SARS of the time needed for preparation once certain issues were settled. In this regard we remind you that a fundamental issue arose as to whether CGT on unit trusts would be imposed on the unit trust management company or on the investor. That decision alone implied a fundamental difference in the way LISPs would henceforth have to conduct their business, and the time required to prepare systems. In fact, at the behest of SARS a survey was conducted amongst our members requesting information on the time required for systems development in the case of the two possible outcomes. The decision only became apparent with the publication of the proposed Bill on 12 December 2000.

Proceeding with CGT with an effective date of 1 April 2001 raises the spectre of rendering a significant financial sector industry unable to comply with the law, without any malice on its side and despite its having cooperated to the extent possible.


The way in which taxable capital gains are included in a taxable income of a taxpayer will influence the calculation of that taxpayer's average rate of tax in terms of section 5(10) of the Income Tax Act, (ITA). The average rate is used in the calculation of the tax payable on lump sums from retirement funds. Is it an intended consequence that the rate at which lump sumps from retirement funds are taxed should be influenced by CGT?

A standardised template is needed for annual reporting purposes with reference to Section 70 A of Act. Will separate IT3B’s be required or can they be combined?


8.1 Administrators will need to perform a data clean up, to include all identity and registration numbers of investors, where applicable. Statistics provided will have to be aggregated on a per client basis. This will take a significant period of time to accomplish.

8.2 Will a person who is required to report to the Commissioner in terms of this section be required to report separately on the gain or loss arising on each disposal or will an aggregate gain or loss figure be sufficient? Is it intended that this will be clarified in the prescribed format for the annual return?

8.3 Certain events which are treated as disposals in terms of paragraph 13 of the proposed Eighth Schedule may not be known to the person required to report, e.g., a change in status from resident to non-resident. There should be no liability for failure to report in such instances. Death which is treated as a disposal in terms of paragraph 29 of the proposed Eighth Schedule may only become known to the person required to report some time after the date of death.

8.4 Various persons/institutions will be required to provide returns to the Commissioner. In certain instances more than one party may be required to report on the same disposal, e.g., both an investment manager with a discretionary mandate to manage a client’s investments and a Lisp administering the investments on behalf of the client appear to be required to report on the same disposals. Such duplication could lead to confusion and should be avoided.

8.5 In order to comply with the reporting requirements imposed by the proposed section 70B and in order to provide clients holding unit trust investments with sufficient information to calculate any capital gains or losses, Lisps will need to calculate the base cost of units in a unit trust fund. The provisions of paragraph 23 dealing with the base cost of assets held at valuation date are complex. (See the comments later). At present, the investor is given the opportunity to elect one of a number of options.

8.6 Units are often acquired over a period of time, at different prices. The method of calculation of the base cost of units acquired over a period of time, after the valuation date, should be clarified. (See our comments later).

8.7 We assume that the proposed Section 70B is intended to cover both pure administration agreements (where the investors exercise the investment decisions themselves) and portfolio management agreements (where the investor outsources the investment decisions to another person).

8.8 Lisps will not be able to provide the information required in terms of this section with the use of their existing systems. A substantial and costly amount of systems development work will be required to enable Lisps to generate this information. This development work will have to be completed prior to 1 April 2001 to enable Lisps to comply with this requirement. It is not possible for Lisps to complete this exercise in two months. Lisps will therefore not be in a position to provide either SARS or its clients with the required returns during the first year of assessment that CGT will be applicable. This will render these provisions of the Act ineffective and lead to disrespect of both the fiscus and the law. It is therefore imperative that the implementation date of the Bill be postponed.


The proposed deletion in Section 103(2) appears to contain a typing error — the words "or after" in "before or after the commencement of the Income Tax Act, 1946" appears to have been omitted.

10. AD SECTION 107
The fact that the envisaged regulations were not made available is a serious shortcoming in the Bill. Not only does that create a gap in the understanding of the tax and its implications, but it does nothing to help the workload for SARS and commentators who have to be consulted in a second similar process.


Should a permanent heterosexual relationship be treated differently from a permanent same sex relationship? In terms of the proposed definition a heterosexual partner will only be considered to be a spouse if "married". If a permanent same sex relationship causes a person to be regarded as a spouse, shouldn't the same apply to permanent non-married heterosexual relationships? This is discriminatory and the definition must be amended to recognise the rights of heterosexual couples that choose not to marry.

Paragraph 2 appears to be reasonably easy to circumvent simply by holding sufficient investment assets in the company. To avoid this, a lower percentage than 80% may be called for.

13.1 The higher annual exclusion than that originally proposed is a vast improvement that should ease some of the administrative burden of both SARS and taxpayers. However, an annual taxable gain of R10 000 is still quite low and the administration of the Act can probably be rendered more cost effective by a further increase in this amount. It is the view of LISPA that an amount of between R25 000 and R50 000 would be more appropriate, particularly if indexing is not achieved. It will provide some protection against the possible taxation of capital where the tax rate exceeds the real rate of capital growth.

13.2 Another issue which arises in this regard is the effect of inflation on this amount, if indexing thereof is not allowed. The R10 000 (hopefully increased as suggested) will become obsolete as effective relief within a short space of time, and thus defeat the object for which it is included. This amount should therefore either be indexed in the absence of a general indexing provision, or be reviewed and adjusted regularly.

In our view sub-paragraph (b) should be deleted, as the annual exclusion is analogous to an exemption, for example the R3000 annual interest exemption. In principle, it is wrong that a person can only claim a loss to be offset against future capital gains if that loss exceeds the annual exclusion. This would be similar to saying that a person can only claim a section 11(a) interest deduction to the extent that it exceeds the section 10 interest exemption. This provision flouts normal tax principles, and requires significant additional record keeping on the part of an individual, SARS and, particularly, Lisps.


The current paragraph 10 should provide for an inclusion rate for all the policyholder funds of a life insurer, and should explicitly provide for untaxed business to have a zero inclusion rate.

16.1 Although it is understandable that this paragraph is intended to be phrased widely, some of the provisions may be too wide. For example, it is difficult to see how the "creation" of an asset can automatically be viewed as a disposal.

16.2 In many cases "variation" or the "damaging" of an asset should also not be viewed as disposal. We assume that in paragraph 11(1)(f) the asset being disposed of is the option itself and not the asset to which the option relates. If this is not the case, this sub-paragraph would be problematical, for example where a person grants an option to purchase a farm but the option is subsequently not exercised.

16.3 We suggest that the references to "trust" in paragraph 11(2)(c) should read "unit trust".

16.4 We assume that the exclusion in paragraph 11(2)(d) will not apply if borrowing is done via the creation or transfer of a financial instrument. This should be clarified.

17.1 Certain events which are treated as disposals in terms of paragraph 13 of the proposed Eighth Schedule may not be known to the person required to report, e.g., a change in status from resident to non-resident. There should be no liability for failure to report in such instances. Death which is treated as a disposal in terms of paragraph 29 of the proposed Eighth Schedule may only become known to the person required to report some time after the date of death. This reinforces the point made elsewhere that a standard reporting template is required and, if reported in accordance with that, the liability of a service provider should cease.

17.2 LISPA suggests the emigration should not be viewed as a disposal as it will discourage immigrants to bring all their assets onshore, or from declaring these assets tax assets without there really being a reason, forcing the sale of assets to either avoid paying or to generate costs so as to be able to pay.

In terms of paragraph 14(1)(a)(I) it is theoretically possible that tax may become payable on the sale of fixed property before the transfer of the property has taken place. This could cause significant cash flow problems for the seller. It is also contrary to the normal practice followed by SARS in developing tax legislation, which is to ensure that the tax becomes payable once the taxpayer is in possession of the funds to pay the tax.

It appears unfair and illogical that capital gains in respect of these assets should be taxable, but capital losses be disregarded.

20.1 It is understandable that capital losses in respect of things like goodwill, copyright and patents be disregarded, because of the ease with which the value of goodwill can be manipulated. However, if proceeded with, it would then seem fairer also to exclude capital gains in respect of goodwill, copyright and patents, etc.

20.2 In the financial sector competitive advantages very often hinge on the efficacy of computer software. Copyright of such software is of value and taxing only gains, but disallowing losses, is inappropriate. Furthermore, internationally the patent law is being developed to a degree that financial products may be patented. This also implies an incorrect tax situation should gains made on such patents be taxable but losses be disallowed.


It is not clear why the forfeiture of a deposit should be disregarded. In our view the normal anti-avoidance provisions and operation of law is sufficient to prevent manipulation in this context.

22.1 These provisions appear to create unwarranted distortions and huge administrative complexity. We assume that it would not be necessary for a unit portfolio or portfolio administrator to attempt to take straddle transactions into account in its reporting.

22.2 In any event the provision does not appear to fulfil its purpose, because it can be circumvented by merely not making any disposals giving rise to a capital gain within the first 45 days of the tax year.

22.3 It is also unfair to only allow losses to be carried forward to the immediately succeeding tax year. Surely where such losses are capped to the value of any gains in the same year, any excess loss should, in fact, be allowed in the same year of assessment, alternatively, be carried forward indefinitely.

22.4 It is not quite clear whether a straddle asset would also include the equity itself, or only derivative instruments based on the equity. See paragraph 21(3).


23.1 A technical problem which immediately arise in respect of Lisps and their administration of unit trust portfolios is that: the unit trust management company and the Lisp do not necessarily know all the costs that the client incurred in relation to the acquisition or disposal of the asset. Conversely, the management company or Lisp may not necessarily know of all expenditure which has been recovered. The result of this is that the records of the unit trust management company/portfolio administrator may differ from the records of the taxpayer, which will cause difficulties on assessment.

23.2 Secondly, the systems of the unit trust management company and the administration systems of the Lisp do not currently provide for the capturing of the investor's value added tax status. In the retail investment industry value added tax paid by the investor would seldom be allowed as an input tax deduction, but the management company or portfolio administrator is not really in the position to know whether this is the case.

23.3 It is practice in the Lisp industry to repurchase units held by clients in unit trust funds in order to pay ongoing fees due to the Lisp. In terms of the provisions of the draft Bill, these fee repurchases would be treated as disposals giving rise to a capital gains tax event. These ongoing fees do not appear to qualify in terms of paragraph 22(1)(b) as expenditure directly related to the acquisition or disposal of the asset. Repurchases in order to pay administration and other service fees due in respect of an investment should not be treated as a disposal for the purposes of capital gains tax.

23.4 Paragraph 22(2)(b) excludes from the Base Cost of an asset so much of the cost of an asset that has not been paid and is not due and payable in the year of assessment in which the asset is disposed of. There is no provision that permits any such amount, that is in fact paid subsequent to such year of assessment, to be recognised in the calculation of a capital loss or gain. It is recommended that a provision to this effect be introduced.

24.1 This paragraph is the most problematic of all the provisions of the draft bill. The vast number of possible permutations will, at best, make this very difficult and expensive to administer and will require a very long lead time to put the necessary systems in place. Even then it would in some cases not be possible to achieve the result required by the draft bill.

In previous discussions with representatives from SARS and the Department of Finance LISPA was led to believe that it would be acceptable for a Lisp to use the weighted average purchase price of financial instruments as its base cost, or to value them on the "first in first out" basis. Administratively the use of an average purchase price would be much more convenient and easier to administer as our database and system already keeps much of the data that would be needed. Changing the system to report on capital gains on this basis would be a significant, but achievable, exercise. Based on the previously published discussion document we assumed that the historical purchase price of the financial instruments would for capital gains tax purposes be reset to the 1 April value, or possibly the higher of the 1 April value or the historical purchase price. Once again, this would be a significant but achievable exercise.

In contrast to this, the current proposal requires major database changes, access to data that may not be available, very complex calculation engines and a feedback mechanism to deal with the taxpayer's elections. In view of the limited time available, we have not yet done a system impact analysis, but it would definitely be impossible to finish such system changes in time for the proposed 1 April implementation date. It must be borne in mind that these systems will have to be in place on the implementation date as the first disposal subject to capital gains tax may take place on the very next day.

According to paragraph 23, elections by the taxpayer may play a significant part in the eventual tax result, and the Lisp is not in a position to make that election on the taxpayer's behalf. Requiring the Lisp to obtain input from the taxpayer in this regard would make the system even more complex, make it even more complex to administer and further negatively affect the relationship between compliance and administration costs and the tax take.

Attempting to comply with that which paragraph 23 requires in its present form a Lisp will have to keep track of the following additional data items:

The historical purchase cost of instruments acquired before the valuation date — separately per instrument per annum (the same instrument acquired in different tax years must according to paragraph 23(10) be treated as separate assets)
Ancillary costs incurred before the valuation date — separately per instrument per annum
The valuation date market value of each instrument
Ancillary costs incurred after the valuation date — separately per instrument per annum
The before valuation date holding period of instruments — separately per instrument per annum
The post implementation date holding period of each instrument — separately per annum
The disposal value of each instrument — separately per annum
The taxpayers' tax elections — separately for each instrument per annum
The relationship of these various values separately per instrument per annum
All previous gains per annum
All previous losses per annum

In all this the calculations are further complicated by the fact that they will have to be done on at least a monthly basis when instruments are sold to recover fees, and in some cases on a daily basis where portfolios are rebalanced on a daily basis.

Much of this data is simply not available, for example historical purchase costs are not tracked separately for assets acquired in different tax years, pre-implementation date ancillary costs are not known, post-implementation date ancillary cost may not be known, the holding period before 1 April may not be known (and certainly not separately for assets acquired in different tax years) and the taxpayers' elections may not be available in time.

Simply put, paragraph 23 is not workable in the retail investment services industry.

The number of permutations of these values which could lead to different results are vast, for example there are 17 different permutations just for the relationship between historical purchase cost, valuation date value and disposal cost.

We assume that "the expenditure contemplated in paragraph 22" includes the expenditure in paragraph 22(1)(a), in other words the historical purchase costs and any related expenditure.

24.2 The Bill provides for all assets to be valued either by the market value, 20% of proceeds, or the time-apportionment method.

Where the market value is used, sections 23(8)(a) and 24 provide that the market value of a financial instrument listed on a recognised exchange is to be the average of the last price quoted in respect of that financial instrument on the recognised exchange on each of the five days of trading preceding the valuation date.

As unit trusts are not traded on an exchange, the above does not apply to them. It is arguable that the same principles should apply to unit trusts in this regard as to listed shares. However, most unit trust portfolios declare a distribution on 31 March of each year. The price of the unit will be inflated on this date, as the price is "cum distribution". On 1 April 2001 the price will reduce by the value of the distribution as at this date the price is "ex distribution". This will mean that the average price over the 5 days before valuation date will be artificially inflated by the dividend.

24.3 The provisions of paragraph 23 dealing with the base cost of assets held at valuation date are complex. A Lisp is required to report to the Commissioner in terms of the proposed section 70B and should be entitled to use the market value (repurchase price) of the unit trust at the valuation date as the base cost.

24.4 Units are often acquired over a period of time, at different prices. The method of calculation of the base cost of units acquired over a period of time, after the valuation date, should be clarified. (Our preferred method of calculation would be to calculate a weighted average purchase price in respect of such units.)

24.5 The full scope of the provisions relative to the valuation of Units in a Unit Trust Portfolio will only be available once the Regulations referred to in Paragraph 23 are published. Meaningful comment in respect of Paragraph 23 can therefore only be provided once these Regulations have been made available for comment.
24.6 Paragraph 23(1) does not permit the inclusion of paragraph 22 costs incurred prior to the valuation date in the base cost of an asset. There appears to be no justification for this distinction. The content of Paragraphs 23(9) and 23(10) appear to recognise such expenses incurred prior to the valuation date. The provisions are therefore inconsistent.

24.7 We assume that where paragraph 23(2) refers to "the expenditure contemplated in paragraph 22" it includes the expenditure in paragraph 22(1)(a), in other words the historical purchase costs and any related expenditure.

24.8 The wording "Where a person has adopted the market value of an asset" used in subparagraph (5) is confusing, as it appears that sub-paragraph 5 will only be applicable if sub-paragraph 4 is applicable. Sub-paragraphs 2 and 5 cannot be applicable simultaneously, as sub-paragraph 2 is only applicable where the disposal proceeds exceed the paragraph 22 expenditure while sub-paragraph 5 is applicable where the disposal proceeds are lower than paragraph 22 expenditure. It is therefore difficult to see how paragraph 23(5) can be applicable if the person has elected in terms of sub-paragraph 2(a). We assume that sub-paragraph 5 is intended to apply once sub-paragraph 4 has become applicable.

24.9 Paragraph 23(8) should probably be extended to include unit trust prices, which are independently valued and published on a daily basis.

24.10 In the case of financial instruments, paragraph 23(10) renders administration and compliance unnecessarily complex without any concomitant benefit. Using an average purchase price would be much better. Bear in mind that there are many unit trust investors that over the years invested amounts of R50 per month (or less) into unit trusts. To treat every such R600 per annum purchase as a separate asset simply does not make sense. It also compounds all the other burdens of the paragraph 23 complexities.

If it is decided to include unit trusts as paragraph 23(8) financial instruments it should be excluded from paragraph 24 as averaging is not necessary in this case — it would be impossible for the individual investor to influence the valuation date pricing of a unit trust.

26.1 In our view this paragraph is unnecessary as paragraph 26 and the general anti-avoidance provisions provide sufficient protection for the fiscus. Although this is clearly not intended it appears that paragraph 27 can be circumvented by merely making two disposals to the connected person.

26.2 Disallowing a capital loss on the disposal of an asset to a connected party is extremely harsh. Not all disposals to connected parties will necessarily be for purposes of avoidance. It is recommended that this provision be amended to recognise bona fide disposals to connected parties. Although paragraph 31 seeks to mitigate the effect of this paragraph it will not provide the appropriate relief in all instances. The short time provided for comment does not permit the drafting of a proposed provision in this regard.

27.1 In terms of paragraph 29, a deceased person is treated as having disposed of his/her assets to the deceased estate at the market value at the date of death. It appears that the deceased estate will be regarded as having disposed of any such asset to the heir, legatee or trustee at that value and such heir, legatee or trustee will be treated as having acquired the asset at that value. In practice there is often a delay between the actual date of death and a Lisp being advised of the death. It is possible that during that period, certain events regarded as disposals may take place, particularly if the assets consist of wrap fund investments. To whom would any resulting capital gains or losses accrue or are they to be disregarded? In addition, the market value of assets transferred to an heir may differ substantially from the market value of such assets at date of death.

27.2 This change from the original discussion document could lead to large cash shortages within the estate of the deceased person. This may compel the executor to realise capital assets that were intended for legatees at a very inconvenient time.

This paragraph should probably also provide for the currency conversion of the valuation date value of pre-existing foreign assets.

It is not clear what happens if these provisions overlap with the straddle assets provisions in paragraph 21.

We assume that paragraph 34(1) provides that the deceased person will be treated as natural person for purposes of this paragraph.

There should be no distinction between life insurance/endowment policies held by the original beneficial owner and such policies which have been ceded outright to a new beneficial owner. In both instances capital gains tax will be levied on the assets held in the policyholder fund in respect of the policy. There is therefore no basis for only disregarding capital gains or losses in the hands of the original beneficial owner and not in the hands of any subsequent owner, as provided for in terms of paragraph 43.

We assume this refers to the disposal of a right to the compensation and that any asset bought with the compensation will be subject to capital gains tax.

We assume the intention of this paragraph is to place these kind of insurance proceeds outside of the capital gains tax net. However, strictly speaking it is not correct to see this as a disposal of an interest in the policy as the person just exercises a right to proceeds in terms of the policy without disposing of the policy.


This paragraph should probably be expanded to also include retirement funds and the untaxed policyholders fund of life insurance companies.