Case No 685/91
IN THE SUPREME COURT OF SOUTH AFRICA (APPELLATE DIVISION) In the matter between:
P A BURGESS Appellant
and
THE COMMISSIONER FOR INLAND REVENUE Respondent
CORAM: CORBETT, CJ, E M GROSSKOPF, VIVIER, KUMLEBEN, NIENABER, JJA
HEARD: 19 May 1993
DELIVERED: 2 Junie 1993
J U D G M E N T
2
E M GROSSKOPF, JA
This is an appeal from the Transvaal Income Tax Special Court, leave having been granted in terms of sec 86A(5) of the Income Tax Act, no 58 of 1962, ("the Act") to appeal to this Court. The point in issue is whether the Commissioner for Inland Revenue was correct in holding that a liability for interest which had accrued was, in the circumstances of the case, not a permissible deduction in terms of sec 11(a) of the Act. The Special Court held in favour of the Commissioner, dismissed the appellant's appeal and confirmed the assessment in which the claim for a deduction was disallowed. The facts are as follows.
The appellant is the managing director of a company doing structural engineering work. He is also a director of various property owning companies, some of which own properties on which the structural engineering business is conducted, and one which owns other property which is let. He
3
derives his income mainly from salaries and director's fees paid by these companies. The appellant owns some shares which are quoted on the stock exchange, and prior to October 1987 he speculated in a small way by buying and selling shares. He also owns some units in a unit trust.
During 1987 the appellant was approached by an insurance broker who told him, in the appellant's words, "over a year period we could make you a considerable amount of money" and that all the appellant had to do was to put up a guarantee of R425 000. The scheme to which the broker was referring, was one which had been initiated by a company called Fenton Investments (Pty) Ltd ("Fenton"). The nature of the scheme is set out and explained in a number of documents which were before the Special Court. In addition evidence was given by Mr M J Gray, a director of Fenton, and the appellant himself. The Special Court did not criticize the evidence of these witnesses in any way and, save in a few respects with which I shall deal later, it was not impugned in argument.
4
From these sources, the following picture emerges.
The essence of the scheme was that money would be borrowed from a bank and invested for a short period (one or two years) in assets such as shares which were expected to appreciate. At the end of the period the assets would be realised, the bank repaid, and the balance pocketed. This type of investment normally carries the risk that the assets might not perform as expected, leaving the investor with no profit and an obligation to repay the loan. In the Fenton scheme, this risk was reduced. The investment would be made, not by purchasing the assets themselves, but by entering into a single-premium pure endowment policy. The insurance company (Lifegro Insurance Limited ("Lifegro")) would then manage the money contributed by the investors by way of premiums in the most advantageous way. Lifegro was considered to be highly skilled in this field. And the key to the scheme was that Lifegro would issue a policy with a surrender value which was guaranteed if the policy was surrendered after a year or
5
after two years. Thus Lifegro undertook to repay the amount invested (i e, the single premium) after a year with a profit of four per cent. This did not, of course, guarantee that investors would make a profit. The loans which the investors made would be at 12½ percent (in the present case), thus leaving the possibility of a loss of 8½ percent. Nevertheless, the possibility of a greater loss was eliminated provided Lifegro was able to meet its commitments.
To cater for the possibility that Lifegro might default, the scheme had a further refinement. The transactions would be entered into, not by the individual investor, but by a partnership en commandite of which he and Fenton would be the members. The individual investor would be the limited partner, and would not be liable for partnership debts except to the extent of the bank guarantee, to which I refer later. Fenton, the disclosed partner, would be liable for the balance of the debts, but, as a company with limited
6
liability and few assets, would not be able to pay them. Thus the bank would effectively carry the risk that Lifegro might not be able to pay.
The bank, for its part, lent the money to the partnership for a year. The partnership paid this amount to the insurance company as a single premium on an insurance policy, and ceded the policy to the bank. Provided the insurance company remained sound the bank would be sure to receive, at the end of the year, the borrowed amount plus four per cent. Since, as already said, the bank's rate of interest was 12½ percent, it required security for the balance. This would be provided by the individual investor in the form of a bank guarantee. The cost of providing the bank guarantee would be the only outlay on the part of the individual investor, and, as stated above, the amount of the guarantee was the most he could lose on the transaction in terms of the partnership deed. In return he would obtain 90% of the profits made by the partnership. Fenton would get the
7
rest.
It was clear that the policy taken out would be a non-standard policy in terms of the Sixth Schedule to the Act, and that the proceeds would fall within the investor's gross income in terms of para (eA) of the definition thereof in sec 1 of the Act. A gain would accordingly be taxable. However, the scheme also provided, so its proposers thought, certain tax advantages. The interest paid to the bank, so they considered, would be tax deductible. Since liability for interest would already accrue during the first year (although interest would only be payable annually in arrear), and no income would be payable before the end of the first year, the liability to pay interest would accrue before the accrual of income from the scheme. This would lead to a tax deferment.
This then was the scheme. One feature which seems unusual was the establishment of a limited partnership. One would not have thought that the parties would have found it necessary to guard against a default on the part of a
8
prominent insurance company. Nevertheless, according to Mr Gray's evidence this was indeed the purpose in interposing the limited partnership between the investor and the financial institutions (i e, the bank and the insurance company). No other plausible reason has been suggested. The existence of the partnership certainly made no difference to the payment of tax. There is accordingly no reason not to accept Mr Gray's evidence on this point.
When the appellant was introduced to this scheme it sounded promising to him. The stock market was booming and there was speculation that he could obtain an 18 or 20 per cent return on a borrowed amount of R5 million. He discussed the project with the insurance broker who had introduced him to it. He also consulted his bank manager. The bank manager "in turn contacted Syfrets and the likes" and came back with a very favourable report on the whole concept. The appellant decided to enter into the scheme. He obtained a bank guarantee for an amount of R425 000 in favour of U A L
9
Merchant Bank ("UAL"). This cost him "something like" R6000. On 27 May 1987 he entered into a partnership with Fenton under the name of Fenton Limited Partnership No. 13. On the same day the rest of the transaction was completed: an amount of R5 million was borrowed from UAL (for a year at a rate of 12½) and paid to Lifegro as a single premium on a policy; the policy was issued and it was ceded to UAL. The commencement date of the policy was 27 May 1987. Lifegro guaranteed that on the first anniversary of the policy its surrender value would be at least R5 200 000. The interest payable to the bank for that year was R625 000. This was secured by Lifegro's guarantee of R200 000 profit on the policy combined with the appellant's bank guarantee for R425 000.
In October 1987 there was a crash on the Johannesburg stock exchange. The value of the fund which Lifegro had created from the money invested by the various Fenton partnerships plummeted. It became clear that Lifegro would not, on the first anniversary of the policy, pay more
10
than the minimum amount guaranteed by it. There were two options open to the appellant. He could either get out at the end of the first year and bear the loss of the amount guaranteed by him, or he could extend the investment for another year. If he took the latter course, he would have had to renegotiate the loan with UAL. A higher rate of interest would have been charged. The total amount of interest would, accordingly, have been higher but Lifegro's guarantee was also somewhat more generous at the end of the second year. Nevertheless, unless the investment performed much better in the second year than it had in the first, the appellant would benefit by cutting his losses and getting out. This was the course he adopted and subsequent events proved him right. The accounts of the partnership for the period ended 29 February 1988 showed a loss of R477 740 in respect of interest payable to UAL. The interest had accrued up to 29 February 1988. This loss was allocated to the appellant.(The appellant's counsel conceded that this was a mistake. In
11
terms of the partnership deed the appellant's liability for partnership debts was limited to R425 000, and if the appeal succeeds the figures will have to be corrected.) In his Income Tax return for the year ended 29 February 1988 the appellant claimed the amount of R477 740 as a deduction. This resulted in a net loss for that year. The Commissioner in his determination of the appellant's taxable income disallowed the deduction in its entirety and issued a second revised assessment accordingly. As I have already stated, the appellant objected to the disallowance and appealed to the Special Court without success. Hence this appeal.
Although the loss in question was suffered by the partnership, it was common cause that in principle the appellant was entitled to deduct his share of the loss from his personal income. See Sacks v Commissioner for Inland Revenue 1946 AD 31. Section 24H of the Act was not yet in force in the relevant tax year and need not be considered. The only question in issue in this appeal is whether the
12
deduction of interest claimed by the appellant is permissible
in terms of the general deduction formula laid down in sec
11(a) of the Act. This provision reads as follows:
"For the purpose of determining the taxable income derived by any person from carrying on any trade within the Republic, there shall be allowed as deductions from the income of such person so derived - (a) expenditure and losses actually incurred in the Republic in the production of the income, provided such expenditure and losses are not of a capital nature;"
The Special Court held that the deduction for
interest had rightly been disallowed because the appellant
had not shown that the expenditure had been incurred in the
production of income derived by him "from carrying on any
trade". The first main issue on appeal was whether the
Special Court was correct in this view.
In terms of sec 1 of the Act " 'trade' includes
every profession, trade, business, employment, calling,
occupation or venture, including the letting of any property
and the use of or the grant of permission to use any patent
... or any design ... or any trade mark ... or any copyright
13
... or any property which is of a similar nature."
Mr Levin, who appeared for the Commissioner in this appeal, argued that there were two reasons why the appellant's activities could not be regarded as the carrying on of a trade. The first was that the appellant's actual purpose in making the investment was to reap the reward flowing from the fiscal advantages. The possibility of the scheme generating a commercial return, he contended, was contemplated, but was merely incidental. In short, he contended that, on the facts properly construed, the appellant did not engage in a trade, the scheme being nothing more than a tax engineering device. The fiscal advantage to which he referred was the tax deferment, which I mentioned earlier, arising from the difference in time between the accrual of the liability to pay interest to UAL and the accrual of the benefits under the policy.
At the outset it should be emphasized that we are not here dealing with an attack by the Commissioner on an
14
alleged tax avoidance scheme in terms of sec 103 of the Act. The sole question before us is whether the appellant was carrying on a trade within the meaning of sec 11(a). And despite the existence of the partnership with Fenton as the managing partner, the parties approached this appeal by concentrating on the appellant's position, and, in so far as it may be relevant, the appellant's state of mind. In my view this was correct. On a proper analysis of the facts the partnership existed only as a part of the structure through which the appellant participated in the scheme. It is the legal effect of his participation that has to be determined.
In support of this part of his argument Mr Levin relied mainly on English decisions. Essentially, he said, the enquiry is whether the transaction "... ought ... when viewed fairly and rationally, to be classed as a trading transaction ..." (see FA & AB Ltd v Lupton (Inspector of Taxes) [1972] AC 634 at 644E]. With this proposition I agree. For the rest the English cases on tax avoidance are not really in point for
15
present purposes, although some of their reasoning may be
relevant. In the recent case of Ensign Tankers (Leasing) Ltd
v Stokes (Inspector of Taxes) [1992] 2 All ER 275 the House
of Lords gave an exposition of the law relating to tax
avoidance schemes after a thorough discussion of earlier
cases. As stated by Lord Goff in the latter case at p 295 d-
e:
"Unacceptable tax avoidance typically involves the creation of complex artificial structures by which, as though by the wave of a magic wand, the taxpayer conjures out of the air a loss, or a gain, or expenditure, or whatever it may be, which otherwise would never have existed. These structures are designed to achieve an adventitious tax benefit for the taxpayer, and in truth are no more than raids on the public funds at the expense of the general body of taxpayers, and as such are unacceptable".
The present case is clearly not of the same type as
those considered in the above mentioned English authorities.
The structures of the Fenton scheme, as I set it out above,
were designed to achieve the commercial results of a short
term gain on the investment of borrowed money with some
limitation on the extent of possible losses. No part of the
16
structures can be described as artificial. Each one was designed for a commercial purpose. The expenses incurred in the cost of the bank guarantee and the interest paid to UAL were actual expenses. The appellant paid them, and had to sell assets to do so. If the expected profit had materialised, the appellant would have had to pay tax on it. Of course, there was also an expected benefit by reason of the tax deferment, but this arose from the very nature of the transaction, viz, that interest became due before any profit was realized. It was not something which was contrived in an artificial way.
Mr Levin's argument on this aspect of the case was based purely on the appellant's supposed purpose in entering into the transaction, viz, on the proposition that, although it was a part of the appellant's purpose to make a profit out of the transaction, his main purpose was to secure the fiscal advantage of a tax deferment. This argument postulates that, but for the appellant's purpose to secure a tax advantage
17
from the scheme, he would have been carrying on a trade.
I do not think that this argument is sound in law, even if the facts supported it. If a taxpayer pursues a course of conduct which, standing on its own, constitutes the carrying on of a trade, he would not, in my view, cease to be carrying on a trade merely because one of his purposes, or even his main purpose, in doing what he does is to obtain some tax advantage. If he carries on a trade, his motive for doing so is irrelevant. See in this regard, Lupton's case, supra, at pp 646C to 647G (Lord Morris of Borth-y-Gest), 655G (Viscount Dilhorne); W T Ramsay Ltd v Inland Revenue Commissioners and Others [1982] AC 300 (HL) at p 323E and the Ensign Tankers case, supra, at pp 291h to 292a. Of course the position might be different if a transaction "is so affected or inspired by fiscal considerations that the shape and character of the transaction is no longer that of a trading transaction" (Lupton's case, supra, at p 647G). That is clearly not the case here. As I have pointed out, the shape
18
and character of the transaction in the present case were inspired entirely by commercial considerations.
Moreover and in any event, the facts do not in my view support Mr Levin's contention. As I have stated above, the Special Court expressed no criticism of the two witnesses who testified in the case, namely Mr Gray and the appellant himself. On the contrary, the Court said: "I place on record that no adverse inference as to their credibility could be drawn from the manner in which the witnesses gave the evidence".
In evidence in chief the appellant stated
categorically that the tax savings or tax benefits did not
play a substantial part in his decision to participate in the
scheme. He was extensively cross-examined on this, but did
not waver. When asked why he entered into the scheme, he
said:
"Well, it looked, as I say, I did react prudently in the sense that I went to the bank manager and tried to get some feedback from that side on a more conservative basis and everything looked very good and obviously the carrot dangled in front of us was
19
the capital gain on it, the growth or the potential growth."
As regards the tax benefit, he said: "It certainly wasn't sold ... to me as a tax benefit". He conceded that the tax benefits were mentioned to him prior to his entering into the scheme, and that these benefits might be significant, but he maintained that it was the prospect of profit which attracted him, not the possibility of a tax deferment.
Mr Gray was to the same effect, although he obviously could not testify as to the appellant's state of mind. He emphasized that the only tangible benefit which could be derived from the tax deferment was that the investor would, for the year or two during which the deferment was effective, be able to invest the money which he would otherwise have had to pay in tax. In the appellant's case he would, according to Mr Gray, have been able to gain a maximum of R6000 in round figures in this way. On the other hand, a
20
profit of 20 per cent on the money invested would have given him a net R310 000. These estimates were not challenged in cross-examination. Counsel's only counter was that the tax deferment was certain while the profit on the investment was uncertain. Mr Gray conceded this.
In argument before us Mr Levin advanced two main reasons why the appellant's evidence in this regard should not be accepted. The first was the emphasis placed in a promotional brochure on the tax advantages accruing under the scheme. The appellant's evidence was that he saw the brochure prior to entering into the scheme, but that he was more impressed by the profit forecasts contained in it than in the tax benefits. There is in my view no reason to doubt his word on this.
The second basis on which Mr Levin tried to cast doubt on the appellant's evidence, was by showing how valuable the tax deferment could have been. In his heads of argument he tried to show that, by investing at 15 persent
21
compound the money he would otherwise have had to pay in tax, the appellant could make a sum of R65 720. It was not explained how the appellant, who was able to borrow money from the bank at 12½ per cent, could obtain 15 per cent from a safe investment. To this calculation the appellant replied, in a set of supplementary heads of argument, with his own calculation which came to a net amount of R5 533. In argument Mr Levin stuck to his figure.
I do not propose analysing the various calculations presented to us. How much the appellant could gain from the tax deferment depends on a number of factors. It would depend firstly on whether he had the money to invest, and on the rate of interest which he could obtain. Then the period for which he could invest the money would depend on how he could juggle his payments of provisional tax. In particular it would depend on his getting the Commissioner's consent in terms of para 19(1)(c) of the Fourth Schedule to the Act to pay provisional tax on less than the basic amount. In the
22
absence of such consent his payment of provisional tax would have had to be based on his most recent assessment. This would reduce the benefit of the tax deferment.
None of these matters was canvassed in evidence. There is no suggestion that the appellant was aware of these rather complicated ways of taking maximum advantage of the tax deferment. In fact, what evidence there is, indicates the contrary. It appears from the appellant's original tax assessment for the 1988 tax year that he had paid his provisional tax on his basic amount. One must therefore assume that he did not receive the Commissioner's consent to pay tax on his expected income (which would in effect be consent not to pay provisional tax in that year at all because he would be incurring a net loss). This payment of provisional tax would, by itself, affect the theoretical calculations put before us, since it was made out of money which, in terms of the respondent's calculations, should have been invested at 15 per cent. Moreover it tends to confirm
23
the appellant's evidence that the tax deferment was not his main reason for entering into the scheme.
In my view the appellant's evidence that his main purpose in making the investment was to make a profit from it has not been impaired in any way, and must stand.
I turn now to the second ground advanced by Mr Levin in support of the proposition that the appellant was not carrying on a trade. This argument presupposes that the appellant's main purpose in entering into the scheme was to earn a profit. Nevertheless, it is contended, "an 'investment' of the nature in question will not, as a matter of law, and on the common cause facts amount ... to the carrying out of a trade" (I quote from the heads of argument).
It is well-established that the definition of trade, which I have quoted above, should be given a wide interpretation. In I T C 770 (1954), 19 SATC 216 at p. 217, Dowling J said, dealing with the similar definition of
24
"trade" in Act 31 of 1941, that it was "obviously intended to embrace every profitable activity and ... I think should be given the widest possible interpretation."
In the present case the appellant argued that its participation in the Fenton scheme amounted to a "venture" which is included in the definition of "trade". In I T C 368, (1939) 9 SATC 211 at p 212, "venture" is defined as "a transaction in which a person risks something with the object of making a profit". This is borne out by dictionary definitions. Thus The Oxford English Dictionary (2nd ed) gives as the appropriate definition "an enterprise of a business nature in which there is considerable risk of loss as well as chance of gain; a commercial speculation." Webster's Third New International Dictionary gives "a business enterprise of speculative nature." See also the definitions quoted in I T C 1476 (1990), 52 SATC 141 at p 148. In the Afrikaans text of the Act sec 11 speaks of a "bedryf" which is defined to include, inter alia, "elke
25
professie, handelsaak, besigheid, diens, beroep, vak of onderneming". The word "onderneming", which corresponds to "venture" in the English text, seems in general somewhat wider although it is capable of bearing the same meaning. Thus it is translated in Bosman, van der Merwe and Hiemstra, Tweetalige Woordeboek, as, inter alia, "venture, risky undertaking".
Now in the present case the appellant clearly, in my view, undertook a venture in the above sense. He laid out the money required to obtain a bank guarantee, and risked the amount of the guarantee, in the hope of making a profit. It was a speculative enterprise par excellence.
In the judgment of the Special Court and the argument on behalf of the Commissioner some doubt was cast on whether the appellant himself realized the risks inherent in the scheme. In my view this does not matter. An undertaking does not in my view cease to be a venture merely because the person pursuing it is of a sanguine temperament.
26
In conclusion on this point I must make it clear that although an element of risk is included in the concept of a "venture" in its ordinary meaning, I must not be taken to suggest that a scheme like the present would only constitute a "trade" if it is risky. Whether it would or not would depend on its own facts. If there is no risk involved, it might still be covered by giving an extended meaning to "venture" or by applying the rest of the definition, which is in any event not necessarily exhaustive. See Meyerowitz and Spiro on Income Tax, para 610. Thus in argument Mr Levin accepted that a person who borrowed money at a low rate of interest and invested it at a higher rate, would be engaged in a trade even if his investment was a safe one. For present purposes it is not, however, necessary to pursue this matter further.
Finally Mr Levin argued that the expenditure in the present case was of "a capital nature" and therefore fell outside the terms of sec 11(a) of the Act. The argument, as I
27
understood it, was as follows. An insurance policy is a capital asset and a gain made by investing in one is, in principle, a capital gain. The legislature has, in para (eA) of the definition of "gross income", included gains received or accrued from insurance benefits under non-standard policies. This does not, however, change the essentially capital nature of the policy or of such gains. Therefore, so it is concluded, any expense incurred in obtaining a policy must also be of a capital nature.
There are, in my view, two answers to this contention. First, in terms of para (eA) any gain made by the appellant from the scheme would be part of his gross income. If one assumes that the policy which produces that income is a capital asset, the appellant would have borrowed money to procure the capital asset which produces the income. In I T C 1124 (1969), 31 SATC 53 at p 55-56, a case similar to the present, Trollip J, as President of the Transvaal Income Tax Special Court, held:
"In the present case the interest paid was the
28
recurrent or periodical charge or 'rental' payable for the continued use by the appellant company of the money lent to it. Such interest was not intended or calculated to, nor did it in fact, improve, augment or preserve those aforementioned capital assets, or form part of or add to the cost of acquiring them or enhance their value. Consequently, we do not think that in the circumstances of this case the interest was so closely identified or associated with those capital assets that it must itself be regarded as being of a capital nature."
These findings are in my view equally applicable to the facts of the present case. See also Meyerowitz and Spiro on Income Tax, para 664, Silke on South African Income Tax, 11th ed, para 7.36, De Koker and Urquhart, Income Tax in South Africa, para 10.3, Commissioner for Inland Revenue v Genn & Co (Pty) Ltd 1955 (3) SA 293 (A) at p 300D and Commissioner for Inland Revenue v Drakensburq Garden Hotel (Pty) Ltd 1960 (2) SA 475 (A) at p 480A.
But, in any event, I cannot agree that the insurance policy in the present case constituted a capital asset. The scheme was a short term speculation with borrowed money. The intention was to surrender the policy after a year
29
or two so as to realize the appreciation of its underlying assets. The scheme does not in its nature differ from the speculative purchase of land or shares with the intention of . re-selling at a profit. An asset so held is not a capital asset and a profit made on its realization does not constitute a capital gain. I do not think authority is needed for this conclusion. And the mere fact that the vehicle used for the speculation was an insurance policy cannot, in my view, make any difference.
For all these reasons I consider that the appellant's claim for a deduction in respect of his liability for interest should have been allowed.
The appeal is allowed with costs, including the costs of two counsel. The order of the Special Court is set aside and the second revised assessment is remitted to the Commissioner for Inland Revenue for re-assessment on the basis that a deduction of R425 000 be allowed for the year ended 29 February 1988.
30
E M GROSSKOPF, JA
CORBETT, CJ, VIVIER, JA KUMLEBEN, JA NIENABER, JA Concur