---

THE SUPREME COURT OF APPEAL OF SOUTH AFRICA

JUDGMENT

Reportable

Case no: 239/19

In the matter between

TELKOM SA SOC LIMITED                                                                             APPELLANT

and

THE COMMISSIONER FOR THE SOUTH

AFRICAN REVENUE SERVICE                                                                   RESPONDENT

 

Neutral citation: Telkom SA SOC Limited v The Commissioner for the South African Revenue Service (Case no 239/19) [2020] ZASCA 19 (25 March 2020)

Coram: CACHALIA, SWAIN, MBHA and MOKGOHLOA JJA and KOEN AJA

Heard: 4 March 2020

Delivered: 25 March 2020

Summary: Income Tax Act 58 of 1962 (the Act) – s 24 I – losses or gains caused by foreign exchange fluctuations – proviso to s 24 I(10) – not a self-standing provision for deduction of a commercial loss unconnected to foreign exchange currency differences – Natal Joint Municipal Pension Fund v Endumeni Municipality 2012 (4) SA 593 (SCA) – unitary but not uniform exercise in purposive interpretation of contracts and statutes – application of contra fiscum rule and presumption that statute law not unjust, inequitable or unreasonable.

 

ORDER

 

On appeal from: The Tax Court, Cape Town (Davis J and assessors):

1. The appellant’s appeal in respect of the foreign exchange dispute is dismissed with costs, such costs to include the costs of three counsel.

2. The respondent’s cross-appeal in respect of the cash incentive bonus dispute is upheld with costs, such costs to include the costs of three counsel.

3. The order of the Tax Court is set aside and is replaced with the following order: ‘The appellant’s appeal is upheld in part and the understatement penalties imposed in the appellant’s income tax assessment for the 2012 year of assessment are set aside’.

 

JUDGMENT

 

Swain JA (Cachalia, Mbha and Mokgohloa JJA and Koen AJA concurring)

[1] The origin of the present dispute lies in a disastrous investment made by Telkom International (Pty) Ltd (Telkom International), a wholly-owned subsidiary of the appellant, Telkom SA SOC Limited (Telkom), when it acquired 75 per cent of the issued share capital in Multi-Links Telecommunications Ltd (Multi-Links), a company registered and tax-resident in Nigeria, during May 2007. To make matters worse, Telkom then acquired the remaining 25 per cent shareholding in Multi-Links during a subsequent year of assessment that ended in March 2009.

[2] In addition, Telkom made a number of shareholder loans to Multi-Links, as it required substantial capital from Telkom to become commercially viable, all of which were denominated in US dollar. By October 2011, a total amount of USD 877 022 900.86 had been advanced to Multi-Links, of which USD 346 000 000 was converted into preference share equity, while the remainder of the loans in the amount of USD 531 022 900.86 were outstanding on the loan account.

[3] From 2009 the prospects of Multi-Link repaying these loans appeared to be remote and it was apparent there was little prospect of Telkom resuscitating the business of Multi-Links. Telkom however continued advancing loans to Multi-Links until October 2011, when Telkom and Telkom International disposed of their equity interests in Multi-Links to HIP Oils Topco Ltd, an unrelated third party. As part of the sale Telkom also sold its rights in respect of its loan to Multi-Links, to HIP Oils Topco Ltd, for USD 100. This occurred in Telkom’s 2012 tax year of assessment.

[4] In its audited financial statements for the 2013 financial year, Telkom reflected the realisation of these loans, in the following terms:

‘[i]n determining the taxable income for the Annual Financial Statements ended 31 March 2012, Telkom included a foreign exchange [FX] gain to the value of R247 million on the realisation of the loan.’

However, in its income tax return for the 2012 year of assessment delivered to the respondent, the Commissioner for the South African Revenue Service (the Commissioner), Telkom instead of reflecting the realisation of the loan as a foreign exchange gain, claimed a deduction in the amount of R3 961 295 256 as a foreign exchange loss, in terms of s 24 I of the Income Tax Act 58 of 1962 (the Act). The effect of this was that what would have been reflected as a taxable income of R3.12 billion, with a resultant tax liability of R875 million, was now reflected as a tax loss of R106 billion, with the result that Telkom was due a refund of the provisional tax paid for that year, in the amount of R822 million. The Commissioner therefore issued an additional assessment for the 2012 tax year, disallowing the deduction of R3 961 295 256 and assessing Telkom for tax in the amount of R425 188 643, as a foreign exchange gain in terms of s 24 I of the Act. This issue will hereafter be referred to as the foreign exchange issue.

[5] Telkom also claimed a deduction of R178 788 421 in respect of cash incentive bonuses paid to Velociti (Pty) Ltd (Velociti), pertaining to the connection of initial subscriber contracts for a specific tariff plan, which Velociti made on behalf of Telkom Mobile. The Commissioner, however, only allowed a deduction R42 256 879 and added back R136 531 542, in terms of s 23H(1)(b)(ii) of the Act. This issue will be referred to as the cash incentive bonus issue.

[6] In addition, the Commissioner imposed an understatement penalty in respect of the 2012 year of assessment in an amount of R91 232 665.64, on the grounds that Telkom’s conduct constituted a substantial understatement of its tax liability. In the view of the Commissioner, it was a standard case which warranted a ten per cent penalty in terms of s 223 of the Tax Administration Act 28 of 2011. This issue will be referred to as the understatement penalties issue.

[7] Telkom therefore appealed to the Tax Court, Cape Town (Davis J and assessors).[1] Telkom’s appeal on the foreign exchange issue was dismissed, but its appeal on the cash incentive bonus issue as well as the understatement penalties issue, was upheld. Telkom, with the leave of the Tax Court, appeals to this Court against the Tax Court’s dismissal of its appeal on the foreign exchange issue. The Commissioner cross-appeals against the Tax Court’s decision to uphold Telkom’s appeal on the cash incentive bonus issue, but does not cross-appeal the decision by the Tax Court, on the understatement penalties issue, which is accordingly not an issue in the appeal.

 

The foreign exchange issue

[8] The resolution of the dispute as to the deduction of R3 961 295 256 by Telkom, must be found in the interpretation of the provisions of s 24 I of the Act, which deals with ‘gains or losses on foreign exchange transactions’. However, before dealing with the detailed submissions of Telkom and the Commissioner, as to the correct interpretation to be placed upon this section, the submission by Telkom that its interpretation that resulted in a foreign exchange loss of R3 961 295 256 reflected the commercial reality of the transaction, whereas the interpretation advanced by the Commissioner, that resulted in a foreign exchange gain of R267 421 739, did not, must be considered. It relied upon the following dictum in Natal Joint Municipal Pension Fund v Endumeni Municipality [2012] ZASCA 13; 2012 (4) SA 593 (SCA) para 18:

‘Interpretation is the process of attributing meaning to the words used in a document, be it legislation, some other statutory instrument, or contract, having regard to the context provided by reading the particular provision or provisions in the light of the document as a whole and the circumstances attendant upon its coming into existence. Whatever the nature of the document, consideration must be given to the language used in the light of the ordinary rules of grammar and syntax; the context in which the provision appears; the apparent purpose to which it is directed and the material known to those responsible for its production. Where more than one meaning is possible each possibility must be weighed in the light of all these factors. The process is objective, not subjective. A sensible meaning is to be preferred to one that leads to insensible or unbusinesslike results or undermines the apparent purpose of the document.’

Telkom submitted that the interpretation advanced by the Commissioner produced a result that was removed from commercial reality and was not sensible or businesslike. It also undermined the purpose of s 24 I of the Act and was unjust, inequitable or unreasonable.

[9] The parties were therefore directed to file supplementary heads of argument dealing with the minority judgment in Commissioner for the South African Revenue Services v Daikin Air Conditioning South Africa (Pty) Limited [2018] ZASCA 66, with specific reference to these submissions by Telkom. The parties were also directed to deal with the application of the contra fiscum rule in the interpretation of the section and whether, apart from this rule, any general distinction should be drawn between the interpretation of fiscal statutes and other statutes. The supplementary heads of argument have been of considerable assistance to this Court.

[10] The minority judgment in Daikin questioned what it stated was the conclusion in Endumeni, namely, that no distinction was to be drawn in the interpretation of contracts, statutes and other documents. Daikin concerned the classification and the correct tariff to be applied in respect of ‘window or wall types, self-contained or “split- system”’ air conditioning machines and parts thereof, in terms of Schedule 1 of the Customs and Excise Act 91 of 1964, for customs duty purposes. The majority judgment at para 14, without reference to the decision in Endumeni, stated the following:

‘There is a further consideration. It is well established that a commercially sensible construction should be preferred . . . It appears quite unbusinesslike to differentiate for customs duty purposes, between “split-system” air conditioning machines of which the indoor units do exactly the same work and the outdoor units are exactly the same, simply because the indoor units are placed on ceilings and not on walls.’

[11] In order to properly examine what was stated in the minority judgment in Daikin, it is necessary to quote extensively from the judgment (paras 30-35):

‘In his judgment Van der Merwe JA invokes this dictum when he concludes that the inclusion of indoor units mounted on ceilings leads to the more sensible commercial construction. It thus becomes necessary to examine the dictum from Endumeni as it might apply to this case.

Contrary to Endumeni . . . which, on the authority of KPMG Accountants (SA) v [Securefin] Ltd 2009 (4) SA 399 (SCA), suggests that there is no distinction in the interpretation of contracts, statutes and other documents, we can find nothing in the judgment of Harms DP in KPMG that prevents a drawing of the distinction that we have drawn between the interpretation of legislation and contracts or similar documents. All that Harms DP said at para 39 in KPMG was that “the rules about admissibility of evidence in this regard do not depend on the nature of the document, whether statute, contract or patent”. Self-evidently, the legislative process which culminates in an enactment, and the subsequent interpretation of that enactment, are quite different from the preceding negotiations which lead to the conclusion of a contract and the subsequent interpretation of the contract. It is difficult to see how “commercial sensibility”, alluded to by Van der Merwe JA, can play any role in interpreting a statute. And a statute must apply to all equally – its interpretation cannot be dependent on a particular contextual setting, nor can it vary from one factual matrix to the next. Context is fact-specific and can be applied in the interpretation of contracts and like documents, but not of statutes.

What is required when seeking to ascertain the meaning of legislation is to subject the words used to an engagement, not with speaker meaning, but with the principles and standards that are appropriate to [the] relevant law-making exercise and the subsequent exercise of legal interpretation. In the case of fiscal legislation, an appropriate standard is the contra fiscum rule which is based upon the idea that no tax can be imposed upon a subject of the State without words in legislation clearly evincing an intention to lay a burden on him or her. . .

Recourse to the meaning of the speakers of words used in a statute is not determined in the same fashion as that of words used in a contract. In order to ascertain the intention of the lawmaker, one must have regard to the appropriate principles of law-making. In the instance of the contra fiscum rule, absent unambiguous language, the rule will be decisive in favour of the taxpayer in cases of doubt. The words employed in the statute must be the primary enquiry to consider whether they admit of any doubt or ambiguity. If not, effect must be given thereto, unless glaring absurdity results which the lawmaker could not have contemplated. . . As is correctly pointed out in Lawsa, this approach, laid down in a number of judgments of our courts relating to the interpretation of a legislative enactment, is based upon the literalist-cum- intentionalist view.

In addition there is the compelling consideration that the Interpretation Act 33 of 1957 applies only to legislation. Section 1 reads:

“1. Application of the Act.

The provisions of this Act shall apply to the interpretation of every law (as in this Act defined) in force, at, or after the commencement of this Act in the Republic or in any portion thereof, and to the interpretation of all by-laws, rules, regulations or orders made under the authority of any such law, unless there is something in the language or context of the law, by-law, rule, regulation or order repugnant to such provisions or unless the contrary intention appears therein.”

This distinction reinforces the view that the interpretation of a statute cannot simply be equated [to] that of a contract. Finally s 39 (2) of the Constitution mandate[s] a recourse to the spirit purport and objects of the Bill of Rights in interpreting any legislation.

Applied to the present dispute, at best for the appellant the words employed may be considered to be open to the interpretation for which it argued. But as we have suggested, the application of speaker meaning as determined by the purpose of the provision, the background and production of the document which appellant seeks to call into aid, is not easily applicable to legislative enactments, including a customs tariff.’ (Authorities omitted.)

[12] What is meant by ‘speaker meaning’ and ‘sentence meaning’ is described as follows in a footnote to the judgment:

‘While the words used in the text to be interpreted are to be classified as sentence meaning, speaker meaning is that which can be attributed to the speaker from an examination of the context and the circumstances which gave rise to the existence of the sentence under examination [in the] interpretative process ’

[13] The view of the minority in Daikin, that Endumeni decided that no distinction was to be drawn in the interpretation of contracts, statutes and other documents, was based upon what was stated in footnote 14 in Endumeni, namely:

‘That there is little or no difference between contracts, statutes and other documents emerges from KPMG Chartered Accountants (SA) v Securefin Ltd and Another 2009 (4) SA 399 (SCA) ([2009] 2 All SA 523) para 39.’

The passage referred to in KPMG Chartered Accountants (SA) v Securefin Ltd and Another 2009 (4) SA 399 (SCA) para 39, reads as follows:

‘Third, the rules about admissibility of evidence in this regard do not depend on the nature of the document, whether statute, contract or patent.’

[14] However, as correctly submitted by counsel for Telkom:

(a)  The passage in KPMG, was concerned with a narrower question, namely, in construing meaning do the evidential rules regarding admissibility change, depending on the instrument? The answer was that they do not. That this is so, is clarified by an examination of the relevant passage in the judgment in KPMG para 39, in context: ‘Second, interpretation is a matter of law and not of fact and, accordingly, interpretation is a matter for the court and not for witnesses. . .

Third, the rules about admissibility of evidence in this regard do not depend on the nature of the document, whether statute, contract or patent’ (Emphasis added.)

(b)  Endumeni asserted that the interpretive technique to be utilised in establishing the meaning of words, as between contracts, statutes and other documents, was essentially a unitary exercise in methodology, but did not assert that it was a uniform one. The exercise was unitary in that whatever the nature of the document, consideration had to be given to the language used in the light of the ordinary rules of grammar and syntax; the context in which the provision appeared; the apparent purpose to which it was directed and the material known to those responsible for its production. The exercise was not uniform, because the background to the preparation and production of the particular document, whether contract, statute or other document, had to be considered from the outset.

(c)  Neither KPMG nor Endumeni suggested that it was irrelevant whether particular words were to be construed as part of a contract, statute or other document, because context was all important, regardless of the nature of the document. Endumeni para 18, emphasised this by stating that the interpretation of words used in the document had to take into account, ‘. . . the circumstances attendant upon its coming into existence’.

(d)  As correctly pointed out by the minority in Daikin, the process which culminates in the conclusion of a contract, is quite different from the legislative process which culminates in an enactment. However, the same fundamental interpretive technique is applied, but always allowing for the context, which includes the background to the preparation and production of the particular contract, or statute, or other document in issue. This is the purposive approach to interpretation and not the literalist-cum-intentionalist view espoused by the minority in Daikin.

[15] As correctly submitted by counsel for the Commissioner, it is axiomatic that a statute must apply to all subjects equally and that its interpretation cannot vary from one factual matrix to the next. It is impermissible to apply a particular meaning to legislation, depending upon the factual situation, in which it is sought to be applied. The statement in Endumeni that, ‘. . . a sensible meaning is to be preferred to one that leads to insensible or unbusinesslike results . . .’ meant that in the process of attributing meaning to the words used in legislation (having regard to the words used, the context and the purpose of the legislation) one possible meaning will be preferred over another possible meaning, because the one meaning yields a commercially insensible result, for all subjects and in the appropriate context (for example commercial legislation).

[16] The reference by the minority in Daikin to the provisions of the Interpretation Act 33 of 1957 and s 39(2) of the Constitution, in support of the proposition that a distinction must be drawn between the interpretation of contracts and statutes, only serves to underline the fact that Endumeni did not suggest that it was irrelevant, whether particular words were to be construed as part of a contract or statute.[2]

[17] It must be emphasised that in Endumeni para 19, it was stated that this approach to the interpretation of documents was consistent with the emerging trend in statutory construction, with Endumeni adopting the second of the two possible approaches mentioned by Schreiner JA in Jaga v Dönges NO and Another; Bhana v Dönges NO and Another 1950 (4) SA 653 (A) at 662G-663A, namely that from the outset one considers the context and the language together, with neither predominating over the other. It is important to recall that in Jaga, the correct approach to statutory construction was described in the following terms:

‘Certainly no less important than the oft repeated statement that the words and expressions used in a statute must be interpreted according to their ordinary meaning is the statement that they must be interpreted in the light of their context. But it may be useful to stress two points in relation to the application of this principle. The first is that “the context”, as here used, is not limited to the language of the rest of the statute regarded as throwing light of a dictionary kind on the part to be interpreted. Often of more importance is the matter of the statute, its apparent scope and purpose and, within limits, its background.’

This approach is echoed in the words of Endumeni para 18, namely:

‘The “inevitable point of departure is the language of the provision itself”, read in context and having regard to the purpose of the provision and the background to the preparation and production of the document.’

[18] I turn to consider the contra fiscum rule. In NST Ferrochrome (Pty) Ltd v Commissioner for Inland Revenue 2000 (3) SA 1040 (SCA) para 17, the rule was described in the following terms:

‘An alternative argument advanced on behalf of the appellant was that subpara (d)(iv) was at least reasonably capable of the construction which the appellant sought to place upon it. Accordingly, so it was contended, the contra fiscum rule required that the subparagraph be so construed. Where there is doubt as to the meaning of a statutory provision which imposes a burden, it is well established that the doubt is to be resolved by construing the provision in a way which is more favourable to the subject, provided of course the provision is reasonably capable of that construction . . . But, where any uncertainty in a statutory provision can be resolved by an examination of the language used in its context, there is no rule of interpretation which requires that effect be given to a construction which is found not to be the correct one merely because that construction would be less onerous on the subject.’

[19] C I Miller The Application of a New Approach to Interpreting Fiscal Statutes in South Africa (2016) para 6.4, in a limited-scope dissertation submitted in January 2016 as part fulfilment of the requirements for the degree of Master of Commerce, at the University of Johannesburg, states the following, with which I agree:

‘It is submitted that the contra fiscum rule still applies in South African law and that it would be incorrect to conclude that the contra fiscum rule has no application in the context of an interpretation of a fiscal provision, anti-avoidance or otherwise. The rule is clearly consistent with the values underlying the Constitution. It is conceded that in the modern era of a purposive approach to interpretation, this rule may have a reduced application when compared to the previous era which favoured a strict literal approach to interpretation which more easily appeared to lead to ambiguity. However, to the extent that following analysis, a purposive approach ultimately yields two constructions which are both equally plausible, it is submitted that the contra fiscum rule should apply and the court should ultimately conclude in favour of the taxpayer.’

[20] Counsel for Telkom submitted that the contra fiscum rule should be applied at the outset, as part of the interpretive technique to be utilised in establishing the meaning of words, contained in a fiscal statute. I, however, agree with the submission by counsel for the Commissioner, that the rule should only be invoked, after an interpretational analysis results in an irresoluble ambiguity as to the meaning of the particular provision in the fiscal statute.

[21] As regards the issue of whether any general distinction should be drawn between the interpretation of fiscal statutes and other statutes, the following dictum in Secretary for Inland Revenue v Kirsch 1978 (3) SA 93 (T) at 94D, sets out the correct approach:

‘There is no particular mystique about “tax law”. Ordinary legal concepts and terms are involved and the ordinary principles of interpretation of statutes fall to be applied.’

[22] I turn to consider the submission by Telkom, that the interpretation advanced by the Commissioner produced a result that was unjust, inequitable or unreasonable. The enquiry is not whether this result is produced after the application of the section to the particular facts of this case, but rather whether this result is produced after the application of the fundamental interpretive technique to the section. Prof L M du Plessis writing in 25 Lawsa 2ed at 334, answers in the affirmative, the question as to whether the common-law presumption that statute law is not unjust, inequitable or unreasonable, still has a function in the constitutional era. By reference to Constitutional Court decisions he states the following:

‘These dicta and others after 1994 dealing with just and equitable statutory interpretation, show that the days of the presumption are not numbered: it can and has been used to guide constitutional interpretation and amplify certain of its procedures, and to supplement, facilitate and mediate resort to constitutional values in statutory interpretation. However, if the presumption’s traditional scope of application is considered, it becomes evident that much of the presumption has been subsumed under more specific and clearly articulated provisions of the Constitution guaranteeing rights and just procedures.’

I can accordingly see no reason why the common-law presumption, which may facilitate resort to constitutional values in statutory interpretation, may not be used as a useful aid in purposive statutory interpretation.

[23] I turn to consider the competing submissions by the parties as to the interpretation of the provisions of s 24 I of the Act, which deals with ‘gains or losses on foreign exchange transactions’. The resolution of the dispute as to the deduction of R3 961 295 256 by Telkom is to be found in this section, because the loan to Multi- Links in US dollars, constituted an ‘exchange item’ as defined in s 24 I(1) of the Act, as it was an amount in foreign currency owing and payable to Telkom.

[24] The relevant provisions of s 24 I, as they read at the relevant time, are as follows:

(a) Subsection 24 I(3), provided as follows:

‘In determining the taxable income of any person contemplated in subsection (2), there shall be included in or deducted from the income, as the case may be, of that person -

(a) any exchange difference in respect of an exchange item of or in relation to that person, subject to subsection (10) ’

(b) The relevant portion of s 24 I(10), which is central to a resolution of the dispute, provided as follows:

‘(10) Subject to the provisions of subsection (7 A), no amount shall in terms of this section be included in or deducted from the income of -

(a)  any resident in respect of any exchange difference determined on the translation of an exchange item to which that resident and any company are parties, where that company is –

(i) a connected person in relation to that resident; or

(ii) a controlled foreign company in relation. . . to that resident. . .

(b)  . .

(c) . .

(d) . .

Provided that where that exchange item is realised during any year of assessment, the exchange difference in respect of that exchange item shall be determined by multiplying that exchange item by the difference between the ruling exchange rate on the date on which that exchange item is realised and the ruling exchange rate on transaction date, after taking into account any exchange difference included in or deducted from the income of that person in terms of this section in respect of that exchange item. ’

(c) ‘Exchange difference’ is defined as:

‘. . . the foreign exchange gain or foreign exchange loss in respect of an exchange item during any year of assessment determined by multiplying such exchange item by the difference between -

(a)  the ruling exchange rate on transaction date in respect of such exchange item during that year of assessment, and –

(i) the ruling exchange rate at which such exchange item is realised during that year of assessment; or

(ii) the ruling exchange rate at which such exchange item is translated at the end of that year of assessment; or

(b)  the ruling exchange rate at which such exchange item was translated at the end of the immediately preceding year of assessment or at which it would have been translated had this section been applicable at the end of that immediately preceding year of assessment, and –

(i) the ruling exchange rate at which such exchange item is realised during that year of assessment; or

(ii) the ruling exchange rate at which such exchange item is translated at the end of that year of assessment.’

As correctly observed by the Tax Court, an ‘exchange difference’ is accordingly either a ‘foreign exchange gain’ or a ‘foreign exchange loss’, determined in the manner set out in the definition.

(d) ‘Ruling exchange rate’ is defined as:

‘. . . in relation to an exchange item, where such exchange item is -

(a) a loan or advance or debt in a foreign currency on -

(i)  transaction date, the spot rate on such date;

(ii) the date it is translated, the spot rate on such date; or

(iii) the date it is realised, the spot rate on such date:

Provided that where the rate prescribed in respect of a loan or advance or debt in terms of this definition is the spot rate on the transaction date or the spot rate on the date on which such loan or advance or debt is realised, and any consideration paid or payable or received or receivable in respect of the acquisition or disposal of such loan or advance or debt was determined by applying a rate other than such spot rate on transaction date or date realised, such spot rate shall be deemed to be the acquisition rate or disposal rate, as the case may be.’

(e) ‘Spot rate’ is defined in s 24 I(1) of the Act as:

‘. . . the appropriate quoted exchange rate at a specific time by any authorised dealer in foreign exchange for the delivery of currency.’

(f) ‘Realised’ means, ‘in relation to an exchange item, where such exchange item is -

(a) A loan or advance or debt in any foreign currency, when and to the extent to which payment is received or made in respect of such loan, advance or debt, or when and to the extent to which such loan, advance or debt is settled or disposed of in any other manner’.

(g) ‘Disposal rate’ means, ‘the exchange rate in respect of an exchange item obtained by dividing the amount received or accrued in respect of the disposal of such exchange item by the foreign currency amount in respect of such exchange item.’

(h) ‘Acquisition rate’ means, ‘the exchange rate in respect of an exchange item obtained by dividing the amount of the expenditure incurred for the acquisition of such exchange item by the foreign currency amount in respect of such exchange item.’

(i) ‘Transaction date’ means, ‘in relation to - (a). . .

(b). . .

(c) a loan or advance owing to a person, the date on which the amount payable in respect of such loan or advance was paid to another person or the date on which such loan or advance was acquired by such person in any other manner’.

[25] As correctly pointed out by the Commissioner, the application of the provisions of s 24 I of the Act to the following aspects of the loan by Telkom to Multi-Links, were not in dispute:

(a) The loan in US dollars to Multi-Links constituted an ‘exchange item’, as it was an amount in foreign currency owing and payable to Telkom.

(b) The loan was ‘realised’ in terms of the definition, when Telkom received USD 100 in the 2012 year of assessment, when the loan was settled.

(c) Telkom and Multi-Links were ‘connected parties’ in relation to each other as defined in s 24 I of the Act, because Multi-Links was a foreign company controlled by Telkom. Consequently, in terms of s 24 I of the Act, Telkom was not obliged to include or deduct any amount from its income, which arose from an ‘exchange difference’ ie a ‘foreign exchange gain’ or ‘foreign exchange loss’, calculated in the manner set out above, where such exchange difference was not realised, but only ‘translated’ and restated at the end of each prior year of assessment.

(d) However, when the loan (the exchange item) was realised by Telkom in its sale to HIP Oils Topco Ltd, for USD 100, Telkom was obliged to determine an ‘exchange difference’ in accordance with the proviso to s 24 I(10) of the Act, in the 2012 year of assessment.

[26] The Commissioner submits that the exchange difference in respect of the loan had to be determined by multiplying the loan, by the difference between the ruling exchange rate on the date on which the loan was realised and the ruling exchange rate on the transaction date, being the date when the loan was advanced. The ruling exchange rate of the loan on the transaction date, being the spot rate, was agreed between the parties.

[27] It is the determination of the ruling exchange rate, on the realisation date of the loan, that lies at the heart of the dispute between the parties. Central to the argument of Telkom was that the proviso to the definition of ‘ruling exchange rate’ applied on the facts of this case, with the result that the ‘disposal rate’ was to be used in lieu of the ‘spot rate’, because the ‘disposal rate’ was another ‘rate’, which was used to determine the consideration, payable for the loan. Telkom submitted that the USD 100 received by it as consideration for the disposal of the Multi-Links loan, was obviously not determined by applying the spot rate, defined as an exchange rate quoted by an authorised dealer at a specific time. The spot rate, as defined, on the relevant date, was 7,9600. According to Telkom, if that rate had been applied, the consideration would have been R3 959 520 551 and not R799, being the then equivalent of USD 100.

[28] Telkom submitted that the pertinent question was whether the consideration of USD 100 was determined by applying a ‘rate’. By reference to the Shorter Oxford English Dictionary 6 ed at 2467, it was submitted that the word ‘rate’ is defined as an ‘estimated value or worth’ or ‘estimation, consideration’ or ‘the price paid or charged for a thing or class of things’ or ‘the amount of or of a charge or payment as a proportion of some other amount or as a basis of calculation’. Consequently, the consideration of USD 100, having been agreed upon by the parties to the Multi-Links transaction, fell within any of these meanings and the language of the proviso indicated that the consideration of USD 100 was determined by applying a ‘rate’.

[29] In addition, as pointed out above, Telkom submitted that in the context of the proviso to the definition of ‘ruling exchange rate’, where another ‘rate’ was applied to determine the consideration payable, the ‘disposal rate’ as defined, was to be used in lieu of the ‘spot rate’ as defined. The word ‘rate’ in the proviso was therefore directly linked with the definition of ‘disposal rate’. Consequently, and so the argument went, the immediate context of the word ‘rate’ was accordingly the definition of ‘disposal rate’. The ‘disposal rate’ was determined by dividing the amount received or accrued in respect of the exchange item (expressed in rand), by the foreign currency amount in respect of such exchange item. The context of the word ‘rate’, therefore indicated that the word did not refer to an exchange rate between currencies, but to an agreement as to value or worth.

[30] The Tax Court rejected the argument of Telkom and concluded that it had impermissibly invoked the provision involving exchange rate gains and losses, in order to deduct a commercial loss, which was completely unconnected to foreign exchange currency differences and that s 24 I of the Act, was not a self-standing deduction provision. It based this conclusion upon the following findings:

(a)  The section was introduced to deal with the problem of how to tax gains or losses, caused by fluctuations in the value of the rand, in circumstances where the underlying transaction had been concluded in a foreign currency. It was designed to ensure that amounts, which had to be taken into account in the determination of taxable income, were converted into rands at a defined exchange rate, thus avoiding disputes as to the rand value of what was received, or expended.

(b)  By contrast, the proviso to s 24 I(10) on which the dispute centred, dealt with the difficulty of applying a rate other than the spot rate. Viewed within the purpose of the section, the word ‘rate’ when used in the proviso, meant an exchange rate; namely a rate that reflected the value of the particular currency in question. The purpose of the section was to solve the problem of amounts to be included in, or deducted for tax purposes, where these amounts were denominated in a currency other than the rand. This meaning accorded with the definition of ‘acquisition rate’ which referred specifically to ‘the exchange rate in respect of an exchange item’.

(c)  Consequently, when the section was read as a whole in order to interpret the proviso to a ‘ruling exchange rate’, what the legislature had in mind was an exchange rate as opposed to a discount rate. The section facilitated the conversion of foreign currency into rands, as opposed to serving as a form of a general deduction for a loss which had little to do with exchange rate fluctuations and everything to do with what appellant conceded, was a disastrous investment. The section dealt with losses or gains caused by foreign exchange fluctuations and was not applicable to a ‘business’ loss of the kind incurred by Telkom.

[31] The Commissioner supported the findings of the Tax Court and submitted that for the proviso to the ‘ruling exchange rate’ to apply, Telkom had to demonstrate that the consideration received by it in respect of the disposal of the loan, was determined by applying a rate other than the spot rate, on the realisation date. However, the consideration for the loan of USD 100 was agreed by reference only to the perceived value of the loan, expressed in absolute terms in US dollars. The Commissioner pointed out that Telkom had not suggested that currency exchange ratios played any role in the determination of the price. In addition, no consideration was given to any ratio to be ‘applied’ to the face value of the loan to yield a sale price, as the proviso required. Because the reference in the proviso to a ‘rate other than [the] spot rate’, was a reference to a currency exchange rate, the argument of Telkom was to be rejected.

[32] The Commissioner also submitted that the proviso only envisaged a currency exchange rate, because the proviso formed part of the definition of ‘ruling exchange rate’. In addition, the amount to be included or deducted under s 24 I(3) of the Act, is an ‘exchange difference’, which is defined as ‘the foreign exchange gain or foreign exchange loss in respect of an exchange item’, and s 24 I is titled ‘gains or losses on foreign exchange transactions’. Consequently, the textual context reflects only exchange rates and differences. The ‘exchange difference’ is calculated by multiplying the face value of the exchange item by the difference between two ‘ruling exchange rates’. The rate on the realisation date must be compared with another ruling rate, which can only be the ‘spot rate’ on the transaction date, or the translation date. The spot rate is a currency exchange rate and it would make no sense to compare that rate with a ‘rate’ that does not reflect the relative value of the two currencies in question. Such a comparison could never yield an ‘exchange difference’.

[33] The Commissioner submitted that Telkom’s argument that ‘rate’ can mean an absolute amount or price, so that USD 100 was the ‘rate’ which was applied and that ‘rate’ means ‘an agreement as to value or worth’, falls to be rejected, because in the context of the proviso, ‘rate’ must refer to a basis of comparison, or calculation between two items, for a number of reasons. The proviso requires that the consideration is ‘determined’ by ‘applying’ the rate. The Concise Oxford English Dictionary 12 ed at 390, states that ‘determine’ means to ‘ascertain or establish by research or calculation’ and at 63 that ‘apply’ means to ‘bring into operation or use’. The consideration must therefore be the result of a process of calculation which puts the ‘rate’ into use as a factor, to produce that result and the only type of rate that is able to perform this function, is one which compares two items against one another, such as a currency exchange rate. Consequently, Telkom’s selection of a discounted absolute price to sell the Multi-Links loan, based on its perceived value, cannot on any sensible basis be described as ‘the determination of the consideration by applying a rate’. On such an approach, the ‘consideration’ and the ‘rate’ are the same thing, and there is nothing to ‘determine’ or ‘apply’.

[34] In my view, the argument of Telkom falls to be rejected for the following reasons:

(a)  When the proviso to the definition of a ‘ruling exchange rate’ is interpreted in the context of the section as a whole, the use of the word ‘rate’ means an exchange rate, that reflects the value of a particular currency in question. A currency exchange rate and not a discount rate is contemplated by the proviso.

(b)  The Tax Court correctly concluded that the purpose of s 24 l(10) of the Act was to solve the problem of amounts to be included in, or deducted for tax purposes, where these amounts were denominated in a currency other than the rand. It was designed to ensure that amounts, which had to be taken into account in the determination of taxable income, were converted into rands at a defined exchange rate, thus avoiding disputes as to the rand value of what was received or expended. The section dealt with losses or gains caused by foreign exchange fluctuations and was not applicable to a ‘business’ loss of the kind incurred by Telkom.

(c)  The central argument of Telkom that the USD 100 received by it as a consideration for the disposal of the Multi-Links loan, was determined by applying a ‘rate’, being ‘the price paid or charged for a thing or class of things’ and that this ‘rate’ fell within the definition of the ‘disposal rate’ (to be used in lieu of the ‘spot rate’ as defined), fails to satisfy the requirement in the proviso that the consideration must be ‘determined’ by ‘applying’ the rate. The consideration must be the result of a process of calculation which utilises the ‘rate’ as a factor to produce that result. The only type of rate that was able to perform this function, was one which compared two items against one another, such as a currency exchange rate. It is quite clear that the consideration for the loan of USD 100 was agreed by reference only to the perceived value of the loan. Currency exchange ratios played no role in the determination of the price.

[35] Accordingly, the submission by Telkom that its interpretation of s 24 I of the Act, that resulted in a foreign exchange loss of R3 961 295 256, reflected the commercial reality of the transaction, whereas the interpretation advanced by the Commissioner, that resulted in a foreign exchange gain of R267 421 739, did not, and was not sensible or businesslike, falls to be rejected. The meaning of the relevant portions of the section, interpreted in context, are clear. As correctly pointed out by the Commissioner, Telkom loses sight of the fact that the section is not intended to deal with the tax consequences of commercial losses. Its operation is limited to gains and losses arising out of currency fluctuations. The fact that Telkom realised a foreign exchange gain on disposal of the loan was a product solely of the fluctuation of exchange rates. In a different year of disposal, Telkom may have suffered a foreign exchange loss. As pointed out by the Commissioner the section is agnostic towards the commercial value of the exchange items in which taxpayers choose to invest.

[36] The Commissioner correctly submitted that what was insensible or unbusinesslike was the contention by Telkom that parties could generate a revenue tax deduction, based solely on the deterioration of the quality of foreign currency- denominated debt by applying s 24 I of the Act, that dealt exclusively with gains and losses as a result of exchange rate differences. In addition, as stated in New Adventure Shelf 122 (Pty) Limited v Commissioner, South African Revenue Service [2017] ZASCA 29; 2017 (5) SA 94 (SCA) para 28:

‘. . . even if in certain instances it may seem “unfair” for a taxpayer to pay a tax which is payable under a statutory obligation to do so, there is nothing unjust about it. Payment of tax is what the law prescribes, and tax laws are not always regarded as “fair”. A tax statute must be applied even if in certain circumstances a taxpayer may feel aggrieved at the outcome.’

[37] In addition, Telkom impermissibly sought to interpret the section by reference to the factual setting, in which it was applied. The fundamental principle is that its provisions must apply equally to all, regardless of the circumstances in which the section is applied. The interpretation placed upon the section accords with its purpose and is neither unjust, inequitable nor unreasonable. Nor are its provisions oppressive, as submitted by counsel for the appellant. As there is no ambiguity in the interpretation of the section the contra fiscum rule is not applicable.

[38] In reaching this conclusion I do not overlook two further arguments advanced by Telkom in support of its interpretation of s 24 I of the Act. The first argument is based upon the Explanatory Memorandum, which preceded the insertion of s 24 I into the Act, by s 21 of the Income Tax Act 113 of 1993. The memorandum provided that the section had:

‘. . . the object of treating, for tax purposes, all gains made and losses incurred in respect of foreign exchange transactions in a manner which takes into account as far as possible the principles of fairness, simplicity, economic reality, current tax principles. ’

[39] Telkom points out that the Explanatory Memorandum gives an explanation concerning the proviso in the definition of ‘ruling exchange rate’, as well as an example of its application. The relevant portions relied upon by Telkom are as follows:

‘The ruling exchange rate on the  realisation date is normally the spot rate.

However, when the loan . . . is disposed of on the date of realisation and the consideration . . .

received or receivable in terms of that realisation or disposal, is calculated by using a rate other than the spot rate on that date, then the “ruling exchange rate” is  the “disposal rate”.

The “disposal rate” will, for example, apply when a person disposes of a loan (asset) of $ 10,000 on a date that the spot rate was R3 per dollar, and the value of the asset is therefore R30 000 (R3 x $10,000), to another person for R29 000,00. The disposal rate will thus be R2.90 per dollar (R29 000 / $10 000).’

Telkom then submits that in the present matter, the amount of R29 000 in the example, would be R799 (the equivalent of USD 100), and USD 10 000, would be USD 531 022 901. The calculation accordingly gives a disposal rate of 0,000002, the application of which rate results in a loss of R3 961 295 256.

[40] However, as correctly pointed out by the Commissioner, the example does not reflect the present facts. In the present case, the consideration for the loan was not agreed to in rand, but in US dollar and it cannot be said that the rand price was ‘determined’ by applying a different rate from the spot rate, as it was determined by reference to the loan’s value. In addition, in the example the price agreed was not the result of the application of some discount rate as a result of the devaluation of the loan, as opposed to pure currency value considerations. Telkom has however applied the spot exchange rate to convert the US dollar price to a rand amount, which the parties in the present case did not do in deciding on the price. The consideration for the Multi- Links loan was not R799 but USD 100.

[41] The second argument advanced by Telkom was based upon the introduction of subsection (4) to s 24 I of the Act, with effect from 1 January 2017. Telkom submitted that this amendment, contradicted the conclusion of the Tax Court, that s 24 I was intended to deal with exchange rate fluctuations and not general deductions. It was submitted that it catered for what the Tax Court referred to as ‘a disastrous investment’. The Commissioner however disputed the interpretation placed upon the amendment by Telkom and submitted that the amended subsection, did not allow for the deduction of an amount as a bad debt and did not cater for a commercial loss. It was submitted that the amended subsection merely reduced the burden on an affected taxpayer, by not requiring the taxpayer to also for account for foreign exchange gains, where such a loss was sustained. It had no bearing on how the commercial loss, as opposed to the foreign currency aspects, was to be dealt with in terms of the Act.

[42] In Patel v Minister of the Interior and Another 1955 (2) SA 485 (A) 493 A – D, the following was stated;

‘There is authority for the view that Acts of Parliament, without having been passed for the express purpose of explaining previous Acts, may nevertheless be used as “legislative declarations” or “Parliamentary expositions” of the meaning of such Acts… It is not surprising that Court’s are cautious in the use of this aid to interpretation, since it is usual for later legislation to amend rather than to declare the meaning of earlier statutes on the same topic. It is, of course, the function of the Courts to expound the true interpretation of the law, including statute law, but where Parliament has clearly shown in a later Act what it meant by an earlier one it seems to me to be not only helpful but even proper to have regard to the later Act in interpreting the earlier.’

The competing submissions by the parties as to the correct interpretation to be placed upon the amendment to subsection (4) of s 24 I of the Act, precludes a finding that Parliament ‘has clearly shown in a later Act what it meant by an earlier one’. This is particularly so, as the introduction of the subsection was intended ‘to amend rather than to declare the meaning’, of the subsection in issue.

[43] The Tax Court therefore correctly dismissed the appeal of Telkom against the additional assessment issued by the Commissioner, on the basis that Telkom ‘invoked the provision involving exchange rate gains and losses in order to deduct a commercial loss which was completely unconnected to foreign exchange currency differences’.

 

The Cash Incentive Bonus Issue.

[44] Telkom made a ‘cash incentive bonus’ payment of R178 788 421 to Velociti in the 2012 year of assessment. The Commissioner allowed as a deduction only R42 256 879, by invoking s 23H(1) of the Act. The issue was whether the Commissioner was entitled to apply the section to limit the deduction in the year of accrual, with the result that the balance paid was spread out over a number of tax years. Telkom successfully appealed against this decision to the Tax Court and the Commissioner now cross-appeals against that finding and seeks to have the assessment confirmed.

[45] The facts which are common cause are as follows:

(a)  Telkom paid cash incentive bonuses to dealers on the connection of the initial subscriber contract in respect of a special tariff plan.

(b)  The amount of R178 788 421 related to connections that Velociti made to Telkom Mobile, which Telkom contends were cash incentive bonuses, for every subscription which Velociti made on behalf of Telkom Mobile.

(c)  Of the R178 788 421 claimed as a deduction by Telkom, the Commissioner added back R136 531 542 in terms of s 23H(1)(b)(ii) of the Act.

[46] The Tax Court, in upholding the appeal, made the following findings:

(a)  The benefit that was attached to the expenditure was the conclusion of the contract with the customer in question.

(b)  Velociti rendered all the services which it was obliged to do in terms of the incentive letters and for which the payment of R178 788 421 was made.

(c)  As a result, there was no basis to add back and disallow R136 531 542 of the cash incentive bonus expenditure by the application of s 23H in the 2012 year of assessment.

[47] The relevant portions of s 23(H)(1), as it read at the relevant time, were as follows:

‘Where any person has during any year of assessment actually incurred any expenditure (other than expenditure incurred in respect of the acquisition of any trading stock) –

(a) which is allowable as a deduction in terms of the provisions of section 11 (a). . .; and

(b) . . . in respect of –

. . .

(ii) any other benefit, the period to which the expenditure relates extends beyond such year of assessment, the amount of the expenditure which shall be allowable as a deduction in terms of such section in the said year and any subsequent year of assessment, shall be limited to, in the case of expenditure incurred in respect of –

. . .

(iii) any other benefit to which such expenditure relates, an amount which bears to the total amount of such expenditure the same ratio as the number of months in such year during which such person will enjoy such benefit bears to the total number of months during which such person will enjoy such benefit or where the period of such benefit is not determinable, such period over which the benefit is likely to be enjoyed:

Provided that the provisions of this section shall not apply –

(aa) Where all the goods or services are to be supplied or rendered within six months after the end of the year of assessment during which the expenditure was incurred, or such person will have the full enjoyment of such benefit in respect of which the expenditure was incurred within such period, unless the expenditure is allowable as a deduction in terms of section 11D (2); or ’

[48] Telkom submitted that from the wording of s 23H(1) of the Act, it was clear that the benefit to which the expenditure related, should extend beyond the relevant year of assessment. This required an enquiry into that to which the expenditure ‘relates’. Relying upon the Shorter Oxford Dictionary 6 ed at 2518, Telkom submitted that the relevant meanings of the word ‘relate’, were as follows; ‘bring (a thing or person) into relation with; establish a connection between’ and ‘have reference to, concern’. In order to establish to what the expenditure ‘relates’, it was necessary according to Telkom, to establish the quid pro quo received as a result of the expenditure having been incurred.

[49] According to Telkom, in terms of the agreement between Telkom and Velociti, a once-off incentive bonus was paid for each new connection (contract) effected by Velociti and the connections were to have been made prior to 30 September 2011, being the date on which the dispensation ended. Therefore, the benefit that was attached to the payment of the cash incentive bonuses related to the new contracts that were concluded. These contracts were concluded before the end of the 2012 year of assessment. Consequently, the benefit did not extend beyond the 2012 year and s 23H of the Act was not applicable. Telkom paid a separate commission for the benefit that it derived from the subscription fees, over the term of the subscription agreement, being 24 months. This ongoing commission was separate and apart from the dealer incentive bonuses payable on the conclusion of the contract. On this basis Telkom submitted that the purpose of the payment of the dealer incentive bonus was to ensure the connection of new customers. The benefit to Telkom was the connection of the new customer. The subscription fees over the term of the subscription agreement was a separate benefit, in respect of which Telkom paid a separate commission, for that benefit.

[50] The Commissioner, however, submitted that the key question was when and how the benefit, in respect of which the expenditure was incurred, was enjoyed. This was because the pleaded dispute turned on, when and how Telkom enjoyed the benefit, received from the cash incentive bonus payment. The Commissioner pleaded that it was the subscription agreement with the client that was the source of the direct benefit to Telkom. The Commissioner also pleaded that the benefit to Telkom, flowed primarily and directly from the service contract, in terms of which the individual customer paid monthly subscription fees. The dealer was a mere facilitator, who brought about the source of the benefits, and the benefits ie the fees, were direct and central to Telkom’s business. It was the agreement concluded between Telkom and the respective dealers which was the indirect source of the benefit.

[51] Telkom, however, pleaded that the full incentive bonus related to connections that Velociti made to Telkom, which was a cash incentive bonus for every subscription the company made, on behalf of Telkom. The benefit was received by Telkom in respect of the amounts paid to Velociti upon the connection of the new subscribers by Velociti. The benefit arising from the obtaining of subscribers fees, over a period of 24 months was described by Telkom, as an indirect benefit.

[52] The Commissioner, in support of the submission that the central issue to be determined was when and how Telkom enjoyed the ‘benefit’, in respect of which the expenditure was incurred, relied upon the following provisions in s 23H:

(a) Section 23H(1)(b)(ii), referred to ‘any other benefit’, where ‘the period to which the expenditure relates extends beyond such year of assessment’.

(b) Section 23H(1)(b)(iii) referred to the ratio based on ‘the number of months during which such person will enjoy such benefit’; and

(c) Paragraph (aa) to the proviso limited the application of the section where the taxpayer ‘will have the full enjoyment of such benefit in respect of which the expenditure was incurred’ within such period ending six months after the year-end.

[53] The Commissioner therefore correctly submitted, that the period to which the expenditure ‘relates’, must be the period during which the benefit is enjoyed. Telkom does not incur the incentive bonus expenditure solely to establish a new connection with a customer. The benefit lies in having a customer who pays subscription fees over the fixed term of the contract. Telkom does not enjoy any benefit immediately upon the conclusion of a new contract. It has nothing to show for it until such time as the connection turns into fee income. That is when Telkom begins to enjoy the true benefits of the cash incentive payments.

[54] The Commissioner therefore correctly submitted that the Tax Court erred, in disregarding the true benefit obtained by Telkom, in the form of the monthly subscriber payments over an anticipated 24 month period. Although the conclusion of the contract benefited Telkom, the enjoyment of that benefit was spread out over the period of the contract, so that the period to which the expenditure related could not be limited to the first year. The Commissioner also correctly submitted that the Tax Court erred in treating as relevant to the application of s 23H, the fact that Velociti had rendered all the services which it was obligated to do in terms of the agreement with Telkom, because this had no bearing upon the central question, being when and how Telkom would enjoy the benefit of the contract.

[55] As regards the submission by Telkom, that it paid a separate ongoing commission to Velociti over the subscription period and that this commission, and not the connection bonus was the quid pro quo for the subscription fees, the Commissioner correctly submitted that the question was whether Telkom enjoyed the benefit of the cash incentive bonus, over the contract period. The fact that another type of payment was made as well, did not render the prior question, irrelevant. In addition, it was not apparent that Velociti had to do anything more to earn its ongoing commission and it was artificial to say that the ongoing commission was any more closely linked to the subscription fees, than the incentive payments.

[56] The Tax Court therefore erred, in concluding that there was no basis to add back and disallow R136 531 542 of the cash incentive bonus expenditure by the application of s 23H, in the 2012 year of assessment.

[57] I grant the following order:

1. The appellant’s appeal in respect of the foreign exchange dispute is dismissed with costs, such costs to include the costs of three counsel.

2. The respondent’s cross-appeal in respect of the cash incentive bonus dispute is upheld with costs, such costs to include the costs of three counsel.

3. The order of the Tax Court is set aside and is replaced with the following order: ‘The appellant’s appeal is upheld in part and the understatement penalties imposed in the appellant’s income tax assessment for the 2012 year of assessment are set aside.’

 

 

____________________

K G B Swain

Judge of Appeal

 

Appearances:

For the Appellant: J J Gauntlett SC QC (with P J J Marais SC and C Louw SC)

Instructed by: MacRobert Attorneys, Brooklyn Lovius Block Inc, Bloemfontein

 

For the Respondent: A R Sholto-Douglas SC (with M W Janisch SC, H Cassim and M Tjiana)

Instructed by: The State Attorney, Johannesburg The State Attorney, Bloemfontein

---