South Africa: Supreme Court of Appeal

You are here:
SAFLII >>
Databases >>
South Africa: Supreme Court of Appeal >>
2018 >>
[2018] ZASCA 73
| Noteup
| LawCite
ST v CT (1224/16) [2018] ZASCA 73; [2018] 3 All SA 408 (SCA); 2018 (5) SA 479 (SCA) (30 May 2018)
Download original files | Links to summary |
THE SUPREME COURT OF APPEAL OF SOUTH AFRICA
JUDGMENT
Reportabl
Case No: 1224/16
In the matter between |
|
|
|
S T |
APPELLANT |
|
|
And |
|
|
|
C T |
RESPONDENT |
Neutral citation: ST v CT (1224/16) [2018] ZASCA 73 (30 May 2018)
Coram: Majiedt, Saldulker and Dambuza JJA and Plasket and Rogers AJJA
Heard: 26 February 2018
Delivered: 30 May 2018
Summary: Divorce – waiver of right to claim maintenance upon dissolution of marriage in antenuptial agreement invalid and unenforceable – accrual – party required to make disclosure of assets in terms of section 7 of Matrimonial Property Act 88 of 1984 must establish proper compliance – party who avers assets excluded from estate for accrual calculation bears burden of proof to establish exclusion and also the nexus between excluded assets and current assets – living annuity not forming part of party’s estate for purposes of accrual calculation.ORDER
On appeal from: Western Cape Division, Cape Town (Weinkove AJ sitting as court of first instance):
[a] The appeal is upheld in part.
[b] Paragraphs 3, 4, 5, 6 and 7 of the order of the high court are set aside and substituted with the following (to avoid confusion, the paragraph numbering in the court a quo’s order is retained) :
‘3. Maintenance:
3.1. The defendant is directed to pay to the plaintiff maintenance as follows:
(a) R18 500 per month for one year as from 1 September 2016;
(b) R13 500 per month for one year as from 1 September 2017;
(c) R8500 per month as from 1 September 2018.
3.2. The obligation to pay maintenance as aforesaid shall endure until the plaintiff’s death or remarriage, whichever occurs first. The maintenance must be paid by way of debit order into such bank account as the plaintiff nominates from time to time and by not later than the first day of each month. The defendant shall be entitled to deduct from the amounts specified in 3.1(a) and 3.1(b) the amounts of maintenance already paid pending the appeal.
3.3. The amounts of maintenance specified in 3.1 above are expressed in nominal terms as at 1 September 2016. The amounts payable as from 1 September 2017 and 1 September 2018 respectively, and as from 1 September of each succeeding year, must be adjusted by the percentage change in the headline inflation rate (also known as the Headline Consumer Price Index) as notified by Statistics SA (or its equivalent) (‘the index’). Such percentage change shall for purposes of convenience be deemed to be equal to the latest index available from Statistics SA on the anniversary date.
4. The accrual in the defendant’s estate is held to be R 8 892 482.
5. The defendant shall pay to the plaintiff half of this amount, minus R70 000 in respect of the plaintiff’s net restitutionary obligation, ie a net amount of R4 376 241, by not later than 1 December 2018. Pending such payment, and as from 5 August 2016, interest shall run on the said net amount at the prescribed rate.
6. The Rondebosch property:
6.1 The plaintiff is ordered to transfer to the defendant her undivided half share in the property situated at 5 Woodlands Road, Rondebosch, Cape Town (the Rondebosch property) free of consideration.
6.2 Such transfer shall be effected as soon as is reasonably practicable after the date of the appeal judgment.
6.3 The reasonable costs of transfer shall be borne in equal shares by the parties.
6.4 Transfer shall be effected by attorneys appointed jointly by the parties, such appointment to be made within one month of the appeal judgment. If the parties cannot agree on the identity of such attorneys within the said one-month period, attorneys appointed by the President of the Law Society of the Cape of Good Hope shall be mandated by the parties to effect transfer.
6.5 The plaintiff shall be entitled to remain in occupation of the Rondebosch property rent-free until one month after the date on which payment of the amount in 5 above is effected. Any agreement for the sale of the property must be subject to this right of occupation.
7. The defendant may collect the movables specified in Exhibit 41 of the record from the Rondebosch property during the one-month period contemplated in 6.5 above.’
[c] The parties shall bear their own costs of appeal.
JUDGMENT
Majiedt JA and Rogers AJA (Saldulker and Dambuza JJA and Plasket AJA concurring):
[1] In this bitterly contested divorce the trial ran for 53 court days and the record is in excess of 8 000 pages. The primary disputes concerned maintenance, accrual and property. The Western Cape Division, Cape Town (Weinkove AJ sitting as court of first instance) (the high court), granted limited leave to appeal to this court in respect of the orders it granted. On petition, this court extended leave to all the issues.
[2] This joint judgment deals with all issues other than the legal basis on which the waiver of maintenance by the respondent is unenforceable. On this limited issue we differ, though we are both agreed that in the event the waiver is unenforceable. The differing routes by which we reach this conclusion do not affect the outcome of the case.
Background
[3] The parties were married to each other in terms of South African law on 17 July 1992 in Hamburg, Germany. They signed an antenuptial contract (the contract) which regulated their marriage. The marriage was out of community of property and the accrual system was included. The appellant, Mr ST, is an experienced advocate, having taken silk in 1989. The respondent, Mrs CT, is of German nationality and is also a lawyer, although she never qualified as such in South Africa. The parties met in Namibia in 1990 while the respondent was visiting there. At that time the appellant, who ran practices in Johannesburg and Namibia, was an Acting Judge in Namibia. In December 1991 the respondent discovered that she was pregnant with the appellant’s child. On 13 February 1992 the respondent broke off the relationship, but the appellant travelled to Hamburg during July 1992 to propose marriage to her.
[4] The appellant arrived in Hamburg armed with a comprehensive antenuptial contract prepared by his Johannesburg attorney, Mr Alick Costa, a family law specialist. The appellant proposed to the respondent and made it clear to her that the contract was an absolute prerequisite for marriage. The contract was in English and the respondent, with the help of a friend, translated it into German. She sought advice on the contract from, amongst others, a German commercial and tax lawyer, a friend who was a non-practising lawyer and worked at a bank, and from her stepfather. It was common cause that neither the respondent nor any one of these advisers had a thorough knowledge of South African matrimonial law.
[5] The respondent requested the appellant in a letter to postpone the marriage to December, but the appellant insisted that they should marry straightaway. In the letter the respondent expressed her concern about her disadvantaged position due to her lack of knowledge of South African law, particularly relating to the custody of children. In the event, the parties got married, with the contract having been signed the day before the wedding. At that time the appellant was 53 years of age and the respondent was 28. It was his second marriage and her first.
[6] The parties’ first child, a son, was born on 4 August 1992. The family lived from September 1992 in Johannesburg where the appellant ran his advocate’s practice. They lived first in Auckland Park and then in Saxonwold. He also kept chambers in Windhoek, Namibia, and his practice extended to Lesotho and Botswana. Another child, a daughter, was born of the marriage on 29 September 1994. The parties moved to Rondebosch, Cape Town, during December 2007. They also acquired a flat in Parkview, Johannesburg, which the appellant used when he was in Johannesburg.
[7] The marriage relationship broke down finally in early 2010 (there had been a breakdown and reconciliation earlier, in 1994). This divorce action was instituted by the respondent in the high court during November 2010. She claimed, amongst others, spousal maintenance, full particulars of the appellant’s current assets and liabilities in terms of s 7 of the Matrimonial Property Act 88 of 1984 (the MPA) and half of the accrual. The appellant counterclaimed for payment, under the actio communi dividundo, of property related expenditure on jointly owned properties and for the return of certain movables. An initial conditional counterclaim for nominal maintenance was abandoned at an early stage of the proceedings. The high court by and large granted the respondent’s claims and dismissed the appellant’s counterclaims. The respondent’s claim for spousal maintenance is contrary to clause 9 of the contract in terms whereof the respondent had waived her claim for maintenance after the dissolution of the marriage (the waiver clause).
The waiver of maintenance
[8] Clause 9 reads as follows:
‘The intended wife accepts the donation in [clauses] 6 and 8 on the conditions stipulated therein and in consideration thereof waives any present or future right whatever that she has or may have to claim maintenance for herself (but excluding maintenance for any dependent child or children born of the intended marriage) on the dissolution of the intended marriage in whatever manner and for whatever reason and regardless of the conduct of the parties.’
The donations in clause 6 and 8 were the half share of a residential property in Twickenham Road, Auckland Park, Johannesburg (the Twickenham property), and the sum of R300 000, payable in three annual instalments from 1992 until 1994, respectively.
[9] The respondent challenged the validity and enforceability of the waiver clause on four broad grounds:
(a) that the clause is per se as a matter of legal principle inconsistent with public policy;
(b) that the effect of the clause is unreasonable, unfair, unjust and thus against public policy;
(c) that the enforcement of clause 9 would be unreasonable and against public policy; and
(d) that the court has an ‘overriding discretion’ to award maintenance, notwithstanding the waiver provisions.
The high court held that the clause is per se invalid and unenforceable. The learned judge also upheld the additional three grounds of the challenge.
[10] For the reasons set out in his separate judgment, Majiedt JA upholds ground (a) and does not find it necessary to consider the other grounds. For the reasons stated in his separate judgment, Rogers AJA upholds ground (d). Either way, we are agreed that in the event clause 9 is not enforceable in the present case.
[11] The high court was thus correct in declaring the waiver to be unenforceable. This necessitates a consideration of the respondent’s maintenance claim and the appellant’s claim for restitution of the donations. But we find it convenient to deal first with the difficult question of the claim for accrual, since that determination will directly affect the question whether any maintenance should be paid to the respondent and, if so, in what amount. The adverse credibility findings made against the appellant, in turn, may have an effect on the question of accrual. We discuss next those adverse credibility findings, together with the favourable credibility findings in respect of the respondent’s testimony.
The high court’s credibility findings
[12] The high court made several adverse credibility findings against the appellant. Many of these findings related to the appellant’s lack of forthrightness and his failure to disclose fully his financial position. It was contended before us on behalf of the appellant that those findings were tantamount to ‘a calculated crusade of character assassination’, contrary to uncontradicted evidence, not supported by any reasoned evaluation, and indicative of the learned trial judge’s patent bias against the appellant. En passant, it is necessary to record that, after the high court had dismissed appellant’s substantive application for postponement well into the trial, a substantive application for his recusal was bought by the appellant. The recusal application suffered the same fate as the application for postponement. We were urged to disregard in totality the learned trial judge’s credibility findings and to evaluate afresh the appellant’s evidence and its probity. Reliance was placed on a number of judgments of this court.[1] In response to the trite principle laid down in Dhlumayo and a long line of subsequent cases that an appellate court has very limited powers to interfere with factual findings made by a trial court, particularly if it depended on credibility findings, we were referred to the passages in Dhlumayo where this court pointed out that the record may reflect factual misdirections by the trial court. This, so counsel for the appellant contended, was such a case.
[13] While the learned trial judge regrettably made numerous improper remarks, sometimes entailing unnecessary personal comments about the appellant, we disagree with the contention that he was patently biased. A careful reading of the record does not bear out that submission. The learned trial judge exhibited considerable judicial patience during a long and difficult trial. The record is replete with numerous interlocutory skirmishes and constant interjections and objections (often without merit) by the appellant’s counsel who appeared for him towards the latter part of the trial (earlier in the trial the appellant had different legal counsel).This necessitated numerous rulings by the trial judge. Having said that, the rulings were most certainly not one-sided throughout. For the reasons that will emerge presently, most of the rulings went against the appellant, correctly so.
[14] The record supports, in large part, the rulings and findings of the learned trial judge. The problem with regard to the constant disruptions by his counsel was exacerbated by the frequent argumentative stance adopted by the appellant under cross-examination. As a seasoned lawyer he often took it upon himself to question the relevance of questions and to argue with the cross-examiner about the basis for certain questions and about the cogency or adequacy of his answers. Apart from being argumentative, the appellant was also evasive and obdurately tendentious at times. The learned trial judge was also correct in finding that in certain specific instances the appellant was mendacious. As will presently appear, these valid criticisms against the appellant regarding his credibility wrongly led the high court to a finding that the appellant had dishonestly concealed the true extent of his estate. It is one thing to find that a litigant’s credibility is questionable, but quite another altogether to ignore the objective facts regarding the true ownership of assets, which is the central enquiry here. Certain of the conclusions reached by the high court are in some instances bereft of any reasoning, and in others not borne out by the proved facts. The learned judge also misdirected himself on the facts and the law in reaching some of his conclusions.
[15] The genesis of the appellant’s unsatisfactory showing as a witness is to be found in the manner in which he conducted the litigation. The learned trial judge described the appellant’s approach to the litigation as a ‘scorched earth policy’. At the outset the appellant adopted an intractable stance on two important legal issues. First, he took the stance that the issue of the validity of the waiver clause and maintenance should be heard separately. An interlocutory application for a declaratory order to that effect was dismissed by Dolamo AJ before the trial commenced. Notwithstanding this ruling, during the trial the appellant frequently reverted to this stance.
[16] Second, in a special plea the appellant contended in limine: (a) that an order for the division of accrual under s 8 of the MPA, can only be made upon application (and not by way of an action for divorce) and (b) that a claim under s 3 of the MPA for half of an accrual can only be brought after dissolution of the marriage and was thus not justiciable in a divorce action. The high court declined to deal with the special plea, electing instead to deal with the matter holistically. Under cross-examination the appellant often reverted to his view of the law.
[17] Whether they were right or wrong, the rulings by Dolamo AJ and Weinkove AJ remained extant until set aside. The appellant’s obdurate persistence in his view of the law in the face of these rulings, must be deprecated. It contributed greatly to the poor quality of his testimony and it was the cause of much of the incessant verbal sparring between counsel throughout the trial. For present purposes it would suffice to point out that this court has held in Brookstein v Brookstein that it is not improper to sue for a decree of divorce and an order in terms of s 3 of the MPA in the same action.[2] Through that decision this court has brought finality to divergent views in some high courts on this particular aspect.
[18] The appellant was correctly criticized by the high court for his recalcitrance in making proper discovery and to furnish adequate replies to requests for further particulars for trial. As far as the accrual claim goes, full and proper disclosure, particularly of his financial affairs, was self-evidently required of the appellant. And s 7 of the MPA unequivocally required of the appellant ‘to furnish full particulars of the value of [his] estate’.
[19] It is of considerable significance that the appellant attested to no fewer than 16 discovery affidavits in this matter. Initially the appellant withheld documents which had a bearing on his assets which were excluded from his estate for purposes of the accrual (the excluded assets). He also failed to discover documents relating to his current assets and liabilities and the trusts and corporate entities in which he held an interest. In particular, no financial statements or any other financial records in respect of the companies and trusts in which it later became clear the appellant had an interest, were disclosed. As was the case with his evidence, the appellant also imposed his view of what was relevant in the discovery process. That much is evident from his first two discovery affidavits. Discovery is not dictated by a litigant’s view of what is relevant – it is a matter for the court, with reference to the pleadings.[3]
[20] Two glaring examples of his failure to make proper disclosure concern, firstly, the appellant’s refusal to produce the liquidation and distribution account (the L & D account) in his late father’s estate and, secondly, his failure to discover documents relating to his so-called Nedbank ‘No 2’ account (the No 2 account). These documents were highly relevant in respect of his alleged excluded assets. It was not in dispute that, upon his father’s death in 1987, the appellant had inherited a substantial share portfolio. According to the appellant, he later realised these shares and acquired his excluded assets with the proceeds. We shall deal with the precise details later. According to the appellant, the No 2 account recorded all the realisations on the sales of assets which existed at the time of the marriage and the purchase of new assets from the proceeds of the sales. It is plain that the L & D account and the No 2 account were of vital importance in an investigation into his excluded assets.
[21] The L & D account was only produced late in the trial, after the respondent had closed her case. The bulk of the statements in respect of the No 2 account were only produced in February 2015, some months after the commencement of the trial. The respondent is correct in her contention that the appellant kept meticulous records. This was particularly evident in his comprehensive archiving of financial documents from the No 2 account and other accounts which had a bearing on his counterclaims. There is a huge gap in statements from the No 2 account for the critical period 1992 until 1998. This was the period during which, on the appellant’s version, there were substantial sales of excluded assets. Self-evidently, the missing statements for that period were of considerable importance.
[22] Equally unsatisfactory were the appellant’s replies to requests for further particulars for trial. In the first request the appellant was asked to make full disclosure of all the assets inside and outside South Africa beneficially owned by him and their market values as well as his liabilities. In his reply on 13 December 2013, the appellant stated that the particulars requested were not strictly necessary for trial preparation, alternatively constituted an interrogatory, or did not arise from the contents of the pleadings, or related to a maintenance claim which had been waived, or related to an accrual claim which could only be adjudicated after a divorce order had been granted. But, significantly, the appellant added that ‘furthermore, [the appellant] has discovered his financial statements as at 29 February 2012, a copy of which will be provided on request and which reflects both his assets and his liabilities. There is no material change therein to date’ (emphasis added).
[23] As will appear later, the reply was untrue. The 2012 financial statements did not reflect the appellant’s shareholding in certain corporate entities at that time. These shareholdings were only disclosed in the appellant’s reply to the second request for particulars for trial, on 17 January 2014. There is considerable merit in the respondent’s contention that, by the time this disclosure was made, the appellant was well aware that the auditors of these entities had been subpoenaed by the respondent and that the information would emerge anyway. Again, as far as the disclosure of his financial affairs in his further particulars for trial are concerned, the appellant had not been forthcoming.
[24] In these two important instances, therefore, the appellant failed to disclose fully important financial information which bore directly on his excluded assets, on new assets realized from the proceeds of the realisation of his excluded assets and thus, overall, on the true value of his estate.
[25] The appellant’s evidence was unsatisfactory in a number of respects. We have already alluded to the fact that he lacked candour, was argumentative and mendacious and evaded questions on the basis that they were irrelevant to the pleaded issues. He was also tendentious at times, particularly on aspects where Dolamo AJ had already ruled against him on his earlier application for a declarator and where the high court had already declined to decide his special plea regarding the accrual claim separately, as outlined above. Reference has already been made to the appellant’s lack of candour in various respects as far as his financial affairs are concerned.
[26] The record bears out the high court’s credibility findings against the appellant. We did not detect any manifestations of bias against the appellant in these findings. Save for some unfortunate and unnecessary personal remarks about the appellant, the learned trial judge cannot be faulted in his credibility findings. In Makate v Vodacom (Pty) Ltd[4] the Constitutional Court, in reaffirming the trite principles outlined in Dhlumayo, quoted the following dictum of Lord Wright in Powell & Wife v Streatham Nursing Home:
‘Not to have seen the witnesses puts appellate judges in a permanent position of disadvantage as against the trial judges, and unless it can be shown that he has failed to use or has palpably misused his advantage, the higher court ought not to take the responsibility of reversing conclusions so arrived at, merely on the result of their own comparisons and criticisms of the witnesses and of their own view of the probabilities of the case.’[5]
[27] While there is no basis on which to interfere with the high court’s credibility findings, that in itself does not, without more, warrant a summary rejection of all of the appellant’s evidence. In particular, as we shall presently demonstrate, absent any acceptable controverting evidence grounded in facts, not conjecture, the appellant’s explanations regarding his financial strategy and its implementation must be accepted where it accords with the objective, established facts as far as the true ownership of assets are concerned.
[28] The high court accepted the respondent’s evidence and made no adverse findings against her credibility. When regard is had to the fact that she had very little knowledge of the appellant’s financial affairs and the true extent of his estate, the respondent’s evidence bears little relevance in the adjudication of this particular issue – the accrual claim. It was contended on behalf of the appellant that the respondent’s evidence ‘was characterised by opportunistic exaggeration coupled with opportunistic understatement, calculated evasion and deflection’. She was also accused of having been vague and having feigned ignorance. We need say no more, given the limited relevance of her evidence, than that the record does not support the attack on her credibility. She was clearly unable to respond meaningfully to the extensive questioning regarding the appellant’s assets, due to her lack of knowledge of his finances. But, importantly, the respondent was unable to adduce any evidence controverting the appellant’s version on this aspect.
Accrual
General introduction
[29] In the contract both parties declared the net value of their respective estates at the commencement of the intended marriage to be nil. It is clear that at the time of the divorce the respondent’s estate was insolvent, taking into account her legal costs. In his plea the appellant denied that his estate had shown a greater accrual than that of the respondent’s. His case was that his current estate was fully attributable to the proceeds realised from his excluded assets. In order to calculate the accrual, the value of the appellant’s net estate must be determined. That value would exclude the assets stipulated as excluded assets in the contract and the current assets acquired from the proceeds of excluded assets. Since the respondent’s net estate has shown no accrual at all, the final computation would only take into account the accrual in the appellant’s estate, if any. The value of the accrual must be determined at the dissolution of the marriage – ie the date of divorce.[6]
[30] The contract contains important provisions relating to the calculation of the accrual. Clause 3 in relevant part reads as follows:
‘3. The provisions of s 5(2) of the [MPA] are hereby excluded from the intended marriage, it being agreed that, in the determination of the accrual of the estate of a spouse, a donation between spouses, including a donation mortis causa, shall be taken into account as part of the estate of the donee . . . ‘[7]
Capital or income received by or accrued to an intended spouse from a third party as an inheritance, legacy or donation were excluded in terms of clause 4.[8]
[31] Both sides conducted the trial and the appeal on the basis that clause 3 of the contract was valid. Since the contrary was not argued, we shall make the same assumption. We simply observe that s 5(1) of the MPA excludes from a spouse’s accrual an inheritance, legacy or donation except in so far as the antenuptial contract provides otherwise, whereas s 5(2) excludes from a spouse’s accrual a donation from the other spouse without adding ‘except in so far as the antenuptial contract provides otherwise’. This may suggest that s 5(2) cannot be excluded by agreement.
[32] In terms of clause 5, assets directly or indirectly owned by an intended spouse at the date of the marriage were excluded. So too were assets acquired from the income derived from such assets and the proceeds from the realisation of such assets.[9] The last part of clause 5 is of considerable significance. It reads:
‘If any funds which would otherwise be included in calculating the accrual of the estate of either spouse be applied by either spouse in the enhancement of any asset which is so excluded, or the payment of any debt which is related to an excluded asset, such funds shall notwithstanding such application be included in the calculation of the accrual of the estate of either spouse.’
The effect thereof is that, for example, any income received by a spouse during the course of the marriage which is used to enhance an excluded asset or to pay a debt relating to such asset, should be included in the calculation of the accrual of the estate of that spouse. Exactly how it should be included will be discussed later.
Duty of disclosure
[33] Section 7 of the MPA sets out the duty which a spouse has to make full disclosure of relevant information when requested to do so by the other spouse. It reads as follows:
‘7 Obligation to furnish particulars of value of estate – When it is necessary to determine the accrual of the estate of a spouse or deceased spouse, that spouse or the executor of the estate of the deceased spouse, as the case may be, shall within a reasonable time at the request of the other spouse or the executor of the estate of the other spouse, as the case may be, furnish full particulars of the value of that estate.’
[34] In MB v DB Lopes J cautioned as follows:
‘. . . litigation is not a game where parties are able to play their cards close to their chest in order to obtain a technical advantage to the prejudice of the other party. This is even more so in matrimonial matters where the lives of the parties have been inextricably bound together. . .’[10]
[35 The duty to make full and frank disclosure in these types of cases has also occupied the attention of the English courts. The applicable legislation contains similar requirements of financial disclosure as ours.[11] In Livesey (formerly Jenkins) v Jenkins[12], Lord Brandon declared, with reference to this duty, that ‘. . . unless the parties make full disclosure of all material facts, the court cannot lawfully or, properly exercise [its] discretion’.[13] That case concerned ancillary orders for financial provision and property adjustment after divorce and the duty to make full disclosure. And, in a more recent case on the same subject, Lord Sumption stated that ‘[t]he proper exercise of these powers calls for a considerable measure of candour by the parties in disclosing their financial affairs. . .’[14] In that case the husband’s conduct was said to have been characterised by ‘persistent obstruction, obfuscation and deceit and a contumelious refusal to comply with rules of court and specific orders’. Mostyn J was right when, with reference to the duty of disclosure, he said that ‘[n]on-disclosure is a bane which strikes at the very integrity of the adjudicative process’.[15]
[36] This court has cautioned that s 7 of the MPA places a clear duty on a spouse to furnish full particulars when called upon to do so. The following dictum of Gorven AJA is apposite:
’39. The attitude of many divorce parties, particularly in relation to money claims where they control the money, can be characterised as “catch me if you can”. These parties set themselves up as immovable objects in the hope that they will wear down the other party. They use every means to do so. They fail to discover properly, fail to provide any particulars of assets within their peculiar knowledge and generally delay and obfuscate in the hope that they will not be “caught” and have to disgorge what is in law due to the other party.
‘40. The conduct of the trial on the accrual claim appears to have been run by the appellant on a “catch me if you can” basis. He clearly failed to comply with the provisions of s 7 of the Act. He delayed providing what were obviously relevant documents until the last minute and then did not discover them. He declined to provide any documents concerning the financial position of Full House Taverns. He did not provide documents which could be used to trace assets derived from the excluded assets.’[16]
A failure by a party to make full disclosure, as required by s 7 of the MPA, may warrant the drawing of an adverse inference where it is reasonable in all the circumstances to do so, that a party has hidden assets.[17]
Onus
[37] The high court held that ‘there was an onus on [the appellant] to show that certain assets were excluded, to identify those assets and to trace those assets to show that they were still there and should remain excluded’. The high court found that, on the evidence, the appellant had failed to discharge this onus with regard to excluded assets.
[38] There are, to our knowledge, only two other decisions where similar findings in respect of this aspect have been made. The high court cited with approval the judgment in AM v JM[18] in that Division. There is also the decision in MB v DB[19] where Lopes J, after citing the judgment of Cloete AJ in AM v JM, held that:
‘It seems to me that in circumstances where the appellant is in possession of all the facts relating to the assets reflected as being excluded in the antenuptial contract, he should bear the onus of demonstrating how they have become converted from time to time, and what their present values are which fall to be excluded from the calculation of his net worth as at the date of the divorce. Although the respondent bears the onus of establishing the monetary value of the share of the accrual in the appellant’s estate to which she is entitled, the appellant is required to show which assets are to be excluded from that calculation, and why.’
[39] This court left the question of the onus open in B v B, stating that it was neither necessary, nor desirable, to decide the issue of the onus.[20] Such a determination appears to be unavoidable in this case. For the reasons that follow, we agree with the high court that the burden of proof was on the appellant with regard to the question whether an asset he owned was an excluded asset. It is trite that a party who claims must prove. In this case the respondent had to prove the appellant’s assets and their value. The respondent, in turn, had to prove that some or all of these assets were excluded. This accords with the analogous case of an insurer who seeks to escape liability on the basis of an exception clause – while an insured must prove an occurrence falling within the primary risk insured against, it is for the insurer to prove that an exception applies.[21]
[40] Furthermore, our law generally does not require a party to prove a negative. Thus, in this instance, the respondent would ordinarily not be required, once she had discharged the onus of proving the appellant’s assets and their value, to prove that any one or more of the assets are not excluded for purposes of accrual. There is the further consideration that a defendant whose defence amounts to a confession and avoidance accepts an onus in that respect.[22] The facts regarding his general financial situation and, in particular, his assets and whether they qualified to be excluded or not, fell within the appellant’s personal knowledge. And he had a statutory duty to make full disclosure. All these factors impel us to the inescapable conclusion that the burden of proof with regard to exclusion fell upon the appellant.
The appellant’s assets
[41] Establishing what exactly the appellant beneficially owned at the time of the divorce, and what his excluded assets and their proceeds are, is a vexed question. The high court adopted a robust approach in this regard. It held that the respondent is insolvent and rejected the valuation of the appellant’s estate by his expert, Mr Greenbaum (whose evidence we shall discuss presently). Consequently, the high court concluded, on the basis of the respondent’s contentions, that the value of the appellant’s estate for purposes of computing the accrual was R22 259 702. The two most contentious findings by the high court are:
(a) that the ‘corporate veil must be lifted’ and that the appellant is the true beneficial owner of certain assets held in companies and in trusts; and
(b) that a Sanlam Glacier living annuity, said to be worth R3 270 368 at the time, must be included as an asset in the appellant’s estate for purposes of calculating the accrual.
[42] Care must be taken not to engage in conjecture when assessing the appellant’s estate. The drawing of inferences must accord with the well established approach laid down by our courts. In civil cases, a respondent who places reliance on circumstantial evidence need not prove that the inference which he or she asks the court to draw is the only reasonable inference – it would suffice if he or she can persuade the court that that inference is the most plausible one amongst a number of possible inferences.[23] It was expressed as follows in Cooper:
‘[T]he court, in drawing inferences from the proved facts, acts on a preponderance of probability. The inference of an intention to prefer is one which is, on a balance of probabilities, the most probable, although not necessarily the only inference to be drawn. In a criminal case, one of the ‘two cardinal rules of logic’ referred to by Watermeyer JA in R v Blom is that the proved facts should be such that they exclude every reasonable inference from them save the one to be drawn. If they do not exclude other reasonable inferences then there must be a doubt whether the inference sought to be drawn is correct. This rule is not applicable in a civil case. If the facts permit of more than one inference, the Court must select the most ‘plausible’ or probable inference. If this favours the litigant on whom the onus rests he is entitled to judgment. If, on the other hand, an inference in favour of both parties is equally possible, the litigant will not have discharged the onus of proof.‘[24]
[43] As stated, the appellant’s failure to fully disclose his financial affairs in the face of a statutory duty to do so, may warrant the drawing of an adverse inference. And he bore the onus regarding his excluded assets and their proceeds. Ultimately, the proper approach in our view is to consider the appellant’s version regarding the ownership of assets and excluded assets against the objective facts, bearing in mind where the burden of proof lies – the respondent had to prove the assets that belonged to him, the appellant had to prove that any assets shown to belong to him were excluded assets.
[44] In an attempt to establish the appellant’s financial position with a view to calculating the accrual, the respondent led the evidence of Ms Danielle Ladopoulos, a chartered accountant. Her brief was to peruse all relevant documentation discovered by the appellant and procured under subpoena by the respondent. She was required to review and analyse these documents to establish the content and value of the appellant’s estate and to determine how and when his assets were acquired and funded. The incomplete and incremental discovery by the appellant resulted in Ms Ladopoulos not being able to properly execute her task. Thus, in her first report Ms Ladopoulos indicated in her conclusion that she was unable to determine the net asset values due to the outstanding documentation and information. She also listed which investigative procedures remained outstanding at that time. In her second report, she again alluded to the continuing difficulty in respect of outstanding information. Based on the available information, she estimated that the appellant’s ‘direct and indirect assets may be valued in excess of R36.7 million’.
[45] The high court placed no reliance on Ms Ladopoulos’ conclusions and estimates. That is understandable, given its speculative nature caused by the paucity of information. Instead, the high court considered the evidence and computations of the appellant’s expert, Mr Hilton Greenbaum, a chartered accountant. In this court the parties made extensive submissions on Mr Greenbaum’s evidence and his conclusions. Mr Greenbaum calculated the appellant’s net estate, before the deduction of excluded assets, to be between R11 million and R12 million. According to him, the total realised from excluded assets, together with retained assets, amounted to R13.92 million. The value of the estate for accrual purposes was R2 076 338 and, once the appellant’s outstanding legal fees due to his attorneys were deducted, Mr Greenbaum concluded that the appellant’s accrual was nil. Ms Ladopoulos and Mr Greenbaum were not expert witnesses in the true sense. The respondent’s counsel described them as ‘compendium witnesses’, which we think is apt. They simply drew conclusions and did calculations from the information available to them.
[46] The high court subjected Mr Greenbaum’s evidence to wide-ranging criticism. It found that, in computing the appellant’s net estate to determine the accrual, the witness had left out material sources of income from the appellant’s practice as a senior advocate which, as stated, extended beyond our borders to Namibia, Lesotho and Botswana. Mr Greenbaum was also criticized for basing his calculations on the assumptions that the proceeds from the excluded assets realised during 1996 to 1998 were not moved offshore or used for purposes other than the acquisition of loan accounts. These assumptions, said the high court, were unsubstantiated by the evidence. Lastly, the high court held that Mr Greenbaum’s valuation of the appellant’s estate excluded certain assets and undervalued others, which meant that these values had to be adjusted.
[47] With regard to the appellant’s evidence concerning his assets and the value of his estate for accrual purposes, the high court had no hesitation in rejecting that evidence. The reasons for the rejection were the appellant’s lack of candour, his inadequate disclosure and constant evasiveness under cross-examination. According to the high court, the appellant had failed to discharge the onus with regard to excluded assets and their proceeds and had failed to establish that he had made full disclosure. And, as stated, the high court held that in order to ascertain the true value of the appellant’s beneficial ownership in certain assets, the corporate veneer of certain trusts and companies had to be removed.
[48] The task of determining the true extent of the appellant’s estate is fraught with difficulty. There is a web of South African and offshore entities which may or may not be beneficially owned, either wholly or in part, by the appellant. To exacerbate matters, some entities underwent repeated name changes and they have different financial year ends. It was contended on behalf of the respondent that this elaborate scheme was ‘a device adopted by the [appellant] to conceal assets and the transfer of funds’. The appellant, on the other hand, confirmed what Mr Greenbaum had submitted, namely that this was merely sophisticated estate planning.
[49] As will be demonstrated presently, on the objective, uncontroverted facts, the appellant’s version cannot, on the probabilities, be rejected altogether, as the high court did. What is clear to us, as was submitted by the respondent’s counsel, that the stark reality is that we will never be able to establish the exact value of the appellant’s estate as at the time of the divorce. We are, nonetheless, duty bound to do the best we can, under the circumstances, to determine that value. As Mostyn J said in NG v SG:
‘If the court concludes that funds have been hidden then it should attempt a realistic and reasonable quantification of those funds, even in the broadest terms. . . . The court must be astute to ensure that a non-discloser should not be able to procure a result from his non-disclosure better than that which would be ordered if the truth were told. If the result is an order that is unfair to the non-discloser it is better that the court should be drawn into making an order that is unfair to the claimant.’[25]
The corporate entities
[50] The record shows that the appellant holds direct and indirect interests in a number of entities, both inside and outside South Africa. It was contended by the respondent that, on the available information, the appellant is the only discernible controlling mind of these entities. As stated, the high court found that these entities were in fact a ‘sham’ and a ‘subterfuge’. It ‘pierced the corporate veil’ and found that the appellant held full beneficial ownership in them. These findings are key to a determination of the extent of the appellant’s estate. The major assets are all ostensibly owned by companies. In our view the correct enquiry is whether the appellant can, on all the facts and circumstances, be said to be the true beneficial owner of the assets. The trial judge conflated the concepts of ‘a sham’ and ‘subterfuge’ on the one hand and ‘piercing the corporate veil’ on the other. That is not the true enquiry which is required here. Moreover, the respondent never pleaded that any of the companies are a ‘book entry’ or a ‘sham’, as the high court found.
[51] The appellant had extensive property and farming assets in Namibia, which he acquired from the proceeds of his inheritance prior to the marriage in 1992. The appellant realised certain of these assets into cash, mostly between 1992 and 1996, in order to fund advances on loan accounts to companies which were related party- entities. According to Mr Greenbaum, assets owned by the appellant and entities controlled by him at the time of the marriage realised approximately R8.8 million from the sales of shares, properties and farms during 1992 to 1996, and a further R2.52 million thereafter. These assets included Richemont, Naspers and Anglo American shares, two farms in Mpumalanga, houses in Windhoek, Namibia, and a farm in Namibia. The primary sources of the realisations after 1996 were a share-block apartment in Brenton-on-Sea and the appellant’s shares in Namibian Factors, which were also excluded assets. The total realisation of assets over the entire period amounted to R11 375 560.
[52] The primary asset and a major bone of contention is the Mont du Toit farm in Wellington in the Western Cape. This wine farm was acquired by Blouvlei Landgoed (Pty) Ltd (Blouvlei) in 1996. Blouvlei has undergone two name changes. In 1998 an adjoining farm, Hawequas, was acquired by Blouvlei. The two farms were farmed as one unit and we refer to them together henceforth as ‘the Wellington farm’. The appellant is the sole director of Blouvlei. A Namibian company, Gamsberg Investments (Pty) Ltd (Gamsberg), acquired by the appellant in 1989, before the marriage, holds 300 400 shares in Blouvlei. Gamsberg holds 100 400 of those shares in trust for the appellant. Tartan Investments (Pty) Ltd (Tartan) holds the remaining 99 600 shares. The appellant testified that he had applied a large part of the cash realised from the sale of his excluded assets as payment for and the development of the Wellington farm. He consequently maintained that the Wellington farm is an excluded asset and that he only holds 25% of Blouvlei through Gamsberg.
[53] The appellant is the sole director of Gamsberg and he holds 150 preference shares in it. Tartan holds 100 ordinary shares in Gamsberg. The Tartan shares have made a significant journey over the years. On 1 April 1989 the shares were issued to the appellant and on 26 June 1992 they were transferred to Windhoek Nominees (Pty) Ltd, which held them in trust for the appellant. On 7 February 1994 the shares were transferred to Zwarwin (Pty) Ltd, which held them in trust for the appellant and, thereafter, on 20 May 2000, the shares were transferred to Mr Klaus Nieft, who also held them in trust for the appellant. On 15 September 2000 the appellant, in his capacity as Gamsberg’s sole director, approved the transfer of the shares to Tartan and the transfer was recorded on a share certificate on the same day. Simultaneously, 150 preference shares were issued to the appellant and registered in his name.
[54] Tartan is itself somewhat of an enigma. It is a company incorporated in the British Virgin Islands. The shares in Tartan are currently held by a blind trust in Monaco, of which the beneficiaries are the appellant’s children from his two marriages. The shares had previously been held by a Guernsey trust, established before the marriage, and were transferred to the Monaco trust in 2005. The beneficiaries of the Guernsey trust, like the Monaco trust, were the appellant’s children. The appellant denied that he personally was a beneficiary of either of the offshore trusts. The respondent has pointed out that the appellant had initially refused to reply to the request for further particulars in respect of Tartan. Thereafter, in his second reply, the appellant said he had no knowledge of details such as the company’s registered address, its shareholders, auditors, bank accounts, assets and liabilities and so forth. The respondent has also emphasized the lack of disclosure or discovery of any financial statements and documents in respect of Tartan.
[55] In 1999 a cellar was built on the Wellington farm where wine was produced. A company, Mont du Toit Kelder (Pty) Ltd (Mont du Toit), is the entity through which the production and marketing of wine was done. Mont du Toit has a 1996 registration number and has also undergone several name changes. The appellant is the sole director of Mont du Toit and its shareholders are Gamsberg (651 shares or 65.1 per cent, of which 151 shares are held in trust for the appellant), Tartan (249 shares or 24.9 per cent) and the appellant (100 shares or 10 per cent).
[56] Clos du Toit (Pty) Ltd, (Clos du Toit) previously known as Investment Facility Company Eight Three Three (Pty) Ltd and then as Mont du Toit Kelder (Pty) Ltd, is a 1998 registered entity whose principal business is the holding of trade marks for the Mont du Toit wines. The appellant is its sole director and the sole registered shareholder is Gamsberg, which holds 100 shares, 26 of which are held in trust for the appellant.
[57] Caledon Street Guest House (Pty) Ltd is a company of which the appellant is the sole director. Its principal business was the operating of a guesthouse in George, but it is presently dormant. In his personal annual financial statements disclosed to the tax authorities, the appellant indicated that he is the sole shareholder of the company, but the share register reflects that Gamsberg is the sole shareholder. Schoongezicht Investments (Pty) Ltd is a Namibian holding and investment company, whose sole shareholder is the appellant’s family trust (a South African trust). Its sole director is Mr Klaus Nieft. As stated, the appellant was previously a beneficiary in the trust, but he was later removed. The rest of the beneficiaries are the appellant’s two children from his marriage with the respondent.
[58] Lastly, as far as the fixed assets are concerned, there is the wine farm, Mas d’ Andrum, near Nimes in the south of France. Its ownership was the subject of a fierce dispute between the parties. The farm, which was purchased in 2004, is registered in the name of a French company, SCEA Mas d’ Andrum, whose shares are held by a company in Luxembourg, Lutsinia SA. The sole director of SCEA Mas d’ Andrum is Mr Berndt Philippe, one of the appellant’s wine consultants from Germany. The shares in Lutsinia SA are held by the Monaco trust.
[59] The farm produces the Mas d’ Andrum and Andrumette wines and, according to the respondent, the Mas d’ Andrum wine label has her family crest on it. The respondent alleged that the appellant had asked her to use her family as a front in respect of ownership, since he did not want it to be known in South Africa that he owned a wine farm in France. She alleged further that the appellant had found the farm while on a family holiday in France. According to her, the appellant was actively involved in the marketing and sales of the wines and in determining the prices for the wines. The appellant had flown in his German winemaking consultants to have a look at the farm before he bought it, said the respondent. The respondent also alleged that the appellant had made available €500 000 towards the purchase price of Mas d’ Andrum, something which the appellant denied categorically. The appellant’s evidence was that he had asked Mr Durham, at that time his contact in respect of the Guernsey trust, whether the trust would make the investment, which the trust agreed to do. He did not deny being involved in the farm’s operations. He did deny−
(a) that he had signed the purchase agreement for the farm;
(b) that he had financed the purchase price;
(c) that he is the beneficial owner or that he has any rights in respect of the farm; and
(d) that he had ever paid any expenses or monies in respect of the farm during the marriage.
[60] We have set out in some detail the ownership structure of the various assets to demonstrate its complexity. We were urged on behalf of the respondent to endorse the findings of the high court, which held that all the evidence compels it to conclude that the appellant, and not the corporate entities, is the true beneficial owner of the Wellington farm. The high court made no such finding in respect of the French farm, correctly so. There is simply not enough information available with regard to Mas d’ Andrum. Counsel for the respondent did not seek to persuade us otherwise. After careful consideration, we have come to the conclusion that the high court’s finding on the ownership of the Wellington farm is incorrect.
Beneficial ownership of Blouvlei, Mont du Toit and Clos du Toit
[61] As stated, the respondent bore the onus to prove the appellant’s beneficial ownership of the assets she averred ought to be included in his estate to determine its accrual and the appellant bore the onus to prove that any such assets were excluded assets.
[62] Although the respondent at times flirted with the contention that the beneficial owner of the farm was not Blouvlei but the appellant, the respondent’s counsel on appeal did not advance that case. Their argument was that the appellant was the beneficial owner of all the shares in Blouvlei. The trial judge’s findings on this part of the case are unclear, though he seems to have concluded that the companies were a facade, that the corporate veil needed to be lifted, and that the appellant was the true owner of the Wellington farm. In the judgment it is stated that ‘[t]he offshore company which he claimed owns part of the land and part of the farm constituted a book entry and a sham’. It is difficult to know how a company can ‘constitute a book entry’. The offshore company which the trial judge had in mind was presumably Tartan. The appellant never claimed that Tartan owned any part of the Wellington farm; he claimed that Tartan owned some of the shares in Blouvlei.
[63] A finding that the appellant is the true owner of the Wellington farm is not compatible with a finding that he is the true owner of all the shares in Blouvlei. If he is the true owner of the Wellington farm, the true ownership of the shares in Blouvlei is irrelevant, because Blouvlei then in truth owned nothing. Conversely, if the appellant was the beneficial owner of all the shares in Blouvlei, there would be every reason for Blouvlei to be the true owner of the Wellington farm. The trial judge’s conclusion about the true ownership of the Wellington farms is really little more than a conclusion. There is no process of reasoning to support it. The high court did not deal with the evidence with a view to determining how it bore on the inherent probabilities. In the absence of proper engagement with the evidence, the high court’s factual conclusion carries no weight.
[64] The appellant’s case was that he is the beneficial owner of 26 per cent of the shares in Blouvlei and Clos du Toit and of 25.1 per cent of the shares in Mont du Toit (though, since there is currently no residual value in the shares of the latter two companies, the position in relation to these companies does not directly affect the calculation of the accrual). In our view, the respondent did not discharge the onus of proving that the appellant was the beneficial owner of 100 per cent of the shares in these companies. It is thus unnecessary to consider whether the respondent was in any event not precluded from advancing this case on the ground that it was not pleaded and on the further ground that Tartan and Gamsberg (whose ownership the respondent sought to impugn) were not joined.
[65] A full burden of proof rested on the respondent to establish the appellant’s 100 per cent beneficial ownership. Well before the parties’ marriage there was in place an offshore trust in Guernsey for the benefit of the appellant’s four children from his first marriage. After the marriage the trust deed was amended to add the children from the second marriage. The appellant’s contact person in relation to the Guernsey trust was Mr Len Durham, formerly a South African attorney. The appellant testified that he was not a beneficiary of the Guernsey trust. When Mr Durham retired in 2005, the assets of the Guernsey trust were transferred to a Monaco trust, the appellant’s new contact being a Mr Bingham, an ex-auditor. The appellant described this as a ‘blind trust’ of which his six children are beneficiaries. In its conventional meaning, a ‘blind trust’ is one in which the beneficiaries have no knowledge of the assets owned by the trust and no say in their administration.
[66] The trial judge regarded the formation of the offshore trust as ‘sinister’. If he said this with reference to a supposed strategy by the appellant of concealing his assets from the respondent, the criticism overlooks the fact that the offshore trust was established before the parties met. The trial judge said that he did not believe the appellant’s evidence that the offshore trust existed for the benefit of his children and not for his own benefit. The trial judge gave no satisfactory reasons for this conclusion. He claimed that the trust deed for the offshore trust reflected the appellant as one of the beneficiaries, but the passage from the record which he cited was dealing with the domestic family trust, not the offshore trust.
[67] The appellant’s financial statements do not show that he ever received a distribution from any trusts. Mr Greenbaum testified that it was not unusual for South African clients not to know the identity of the trustees of offshore trusts. Even if the appellant were a beneficiary of the offshore trust, and even if he did receive income from it, this would be irrelevant at least for purposes of accrual, since clause 4 of the antenuptial contract excludes from the ambit of accrual any benefit, whether by way of income or capital, received by either of the spouses in terms of any trust howsoever created or to be created.
[68] We have already alluded to the appellant’s domestic family trust. It was established in July 1995 with the appellant as first trustee. The family trust held, among other assets, all the shares in the company Schoongezicht. This company must have existed before the marriage because it owned various pre-marital assets such as the Mpumalanga farms, the shares in Namibian Factors and a 5 per cent shareholding in Pitje Chambers. If the appellant previously was the owner of the shares in Schoongezicht, he must have transferred them to the family trust on loan account, the loan account being repaid when Namibian Factors and the Mpumalanga farms were sold.
[69] If Schoongezicht’s new shareholder had been an entity with no apparent connection to the appellant, one might have reason to query why he allowed control of Schoongezicht to pass from himself to such entity and to investigate whether that entity did not in truth continue to hold Schoongezicht for the appellant’s benefit. However, there is no reason for the appellant to have been unwilling to benefit a family trust for the advantage of his children. It is not unusual for wealthy parents to do this. Although the family trust’s ownership of Schoongezicht was not challenged, we mention this because it is germane to one’s assessment of the other corporate structures.
[70] Prior to the appellant’s marriage to the respondent, there existed the company Gamsberg, which owned Groot Gamsberg on which the appellant farmed in Namibia. Here one already had a model in terms of which the land was owned by a company while the appellant personally conducted the farming operations. This was advantageous as the appellant was allowed to deduct his Namibian farming losses from his Namibian professional income. As set out earlier, originally (ie from around 1989) the appellant personally held all the shares in Gamsberg. In June 1992 the shares were registered in the name of a professional nominee but the appellant remained the beneficial owner by way of a declaration of trust. The identity of the professional nominee changed in February 1994 and May 2000 but by way of declarations of trust the appellant continued to be the beneficial owner of all the shares.
[71] The respondent, of course, knew that the appellant farmed on Groot Gamsberg. Any enquiry into the company Gamsberg would have revealed, until September 2000, that the shares were held by a nominee for the benefit of the appellant. If the appellant had wished to conceal his beneficial ownership, he would not have used professional nominees which issued declarations of trust.
[72] Before discussing the change which occurred in the shareholding of Gamsberg in September 2000, we must go back to the middle of 1996 when Blouvlei bought the Wellington farm. The latter company was set up on the basis that 76 per cent of the shares were registered in the name of Gamsberg and 24 per cent in the name of Tartan. As stated, Tartan was an offshore company wholly owned at that time by the Guernsey trust (now by the Monaco trust). As in the case of the family trust’s shareholding of Schoongezicht, there is no reason to be distrustful of the genuineness of Tartan’s shareholding in Blouvlei, since any beneficial opportunity which the appellant made available to Tartan through Blouvlei would advantage his children and reduce his estate duty. As at 1996 the appellant was the beneficial owner of all the shares in Gamsberg, even though they were held for his benefit by a nominee. Until September 2000 the position was thus that Tartan held 24 per cent and Gamsberg 76 per cent of Blouvlei. Because he was the beneficial owner of all the shares in Gamsberg, the appellant indirectly had a 76 per cent stake in Blouvlei.
[73] When the shareholding in Gamsberg changed in September 2000 as described previously, there was no aspect of the parties’ marriage which would, on the probabilities, have spurred the appellant into taking devious action. Up to then, he indirectly held 76 per cent of Blouvlei. The ordinary shares in Gamsberg were probably worth more than a nominal sum in September 2000, but since the shares were being transferred to an entity wholly owned by a trust which existed for the benefit of the appellant’s children, there is nothing suspicious about the arrangement.
[74] The beneficial ownership of Gamsberg by an entity which existed for the benefit of the appellant’s children, coupled with preference shares issued to himself, has all the hallmarks of a conventional estate planning transaction. For as long as he survived, his preference shares ensured that he retained voting control over Gamsberg, and thus indirectly over Blouvlei, but the value would reside with Tartan and thus indirectly with the offshore trust. If the appellant had been the beneficial owner of Gamsberg, there would have been no need for preference shares to be issued to him. The experts in the present case agreed that the appellant’s preference shares in Gamsberg should be valued at their nominal value, namely R150.
[75] In November 2000 Gamsberg executed a declaration of trust that, of its 76 per cent shareholding in Blouvlei, 26 per cent was held for the benefit of the appellant. The appellant’s evidence was that he wanted to ensure that he personally had a sufficient shareholding to block a special resolution. So, by November 2000 Tartan held 24 per cent of Blouvlei, Gamsberg beneficially owned 50 per cent while the appellant beneficially owned 26 per cent.
[76] The same applies to Mont du Toit which was set up in 1996 to conduct the wine-making operations. There seems to have been some delay in establishing the intended shareholding. Originally 249 shares were issued to Tartan and 100 shares to the appellant personally. In January 2000, 651 shares were issued to Gamsberg, bringing the total number of issued shares to 1000. In November 2000 Gamsberg executed a deed of trust declaring that 151 of the shares which it held in Mont du Toit were owned beneficially by the appellant. The result was that by November 2000 Tartan owned 24.9 per cent of Mont du Toit, Gamsberg beneficially owned 50 per cent, while the appellant beneficially owned 25.1 per cent (of which 10 per cent was registered in his name and 15.1 per cent in Gamsberg’s name). In the case of Clos du Toit, which was set up to own the trade marks used in relation to Mont du Toit’s wines, the arrangement was somewhat simpler, as set out earlier.
[77] As stated, until September 2000 Gamsberg was beneficially owned in its entirety for the appellant. There is thus no reason to doubt that until September 2000 Gamsberg was the beneficial owner of 76 per cent of Blouvlei. It would have been unnecessary for Gamsberg to hold the said shares as a nominee for the appellant, because he was in any event the beneficial owner all the shares in Gamsberg. And we see no reason to find that Gamsberg’s shareholdings in Blouvlei underwent a change of character and ownership in or after September 2000. It would be entirely consistent with the appellant’s declared estate planning purposes that Gamsberg should have continued beneficially to own 76 per cent of Blouvlei. Indeed, the estate planning advantages of the structure above Gamsberg would have been pointless and ineffective if, in truth, Gamsberg itself did not own anything beneficially. The same analysis applies to the shareholdings in Mont du Toit and Clos du Toit.
[78] Furthermore, we know that in November 2000 Gamsberg executed deeds of trust to declare that a part but not all of the shares held by it in the three companies belonged to the appellant. If the intention was for all the shares held by Gamsberg in these companies to be held beneficially for the appellant, why did the declarations of trust not say so? If there was a Machiavellian scheme to cheat the respondent, one would have expected there to be no declarations of trust at all, the true arrangements being sealed with a nod and a wink between conspirators. Alternatively, if there was no honour among thieves, one would have expected the appellant to safeguard himself with a declaration that all the shares were held for his benefit. A declaration as to only a part of the shareholding makes no sense, except on the supposition that the remaining shares were beneficially held by Gamsberg.
[79] The appellant explained in his evidence why the activities connected with the Wellington farm were separated among various companies. Because the wines were to be marketed abroad, including in the United States, he wanted the wine-making company, Mont du Toit, to be ring-fenced so that any claims for damages would not impinge on the Wellington farm. The trade marks were housed in Clos du Toit because his first prize would have been for the wines to be marketed under the name ‘Clos du Toit’. The wine authorities, however, would not allow him to use the word ‘Clos’. He accepted this for the time being, having in mind in due course to take up the fight again. Blouvlei was the owner of the land while he, the appellant, conducted the farming operations. The personal conduct of the farming operations was advantageous because he could set off his farming losses against his professional income. He is nevertheless a bona fide farmer. He has farmed in Namibia and South Africa for decades.
[80] Something was made of the fact that three cottages on the Wellington farm are rented out by Mont du Toit rather than Blouvlei, the suggestion being that this demonstrated a failure to respect the separate corporate personalities of the companies. Even if that were so, it would not show that the appellant personally owned all the shares in the companies. When this was put to the appellant, he made the point that, since the indirect shareholdings in Mont du Toit and Blouvlei were practically identical, it did not matter economically to the ultimate shareholders whether the renting was done by Mont du Toit or Blouvlei. The appellant also said that Blouvlei had not incurred the expenditure in improving the cottages.
[81] In similar vein, the appellant was attacked regarding the arrangements between himself and Blouvlei in connection with his use of the land for farming. He paid Blouvlei a nominal annual rent of R25 000. The first criticism was that, in response to a question in the respondent’s request for trial particulars as to whether there was any lease agreement concluded between the appellant and Blouvlei, the appellant replied ‘none’. He explained in evidence that he had answered the question with formal written lease agreements in mind. The nominal rent paid Blouvlei was apparently an arrangement implemented at the suggestion of the auditors so that Blouvlei could maintain its VAT status. No written agreement existed. While the explanation for his reply to the request is not altogether satisfactory, we do not think one can put this down to dishonesty rather than a slipshod and contemptuous response to the respondent’s probings.
[82] Another criticism was that the appellant was only paying a derisory rent for the farm, showing that he was in truth its beneficial owner. This overlooks two important matters. Firstly, Blouvlei has benefitted enormously from the improvements to the farm made by Mont du Toit (in respect of the wine-making facilities) and by the appellant (in respect of the planting of vineyards). If Mont du Toit and the appellant ceased their respective operations and Blouvlei were to sell the farm, it would make a substantial capital profit from expenditure incurred by others. Second, the appellant has not, in consequence of the ‘derisory rent’, been enabled to make substantial personal profit. To date the farming operations have run at a loss. If and when the operations become profitable, it may be more important to establish market-related rentals (though in the case of Mont du Toit, as we have said, it does not really matter because the ultimate shareholdings in Mont du Toit and Blouvlei are practically the same).
[83] At the conclusion of his testimony the appellant asked to be allowed to say one more thing regarding the attack that his companies and business activities were a pretence:
‘They are not under any circumstances. They are substantive arrangements that I have made and I want to underline that these arrangements – I’m 76 years old – these arrangements were made for the benefit of my children, who I regard as the only important thing in my life.’
[84] We thus do not think that the evidence justified the high court’s finding that the appellant is the beneficial owner of all the shares in the three companies. The assessment of the accrual should proceed on the basis that the appellant owns 26 per cent of Blouvlei, 25.1 per cent of Mont du Toit and 26 per cent of Clos du Toit.
[85] We conclude this part of our judgment by returning to the question of the appellant’s unsatisfactory discovery and replies to requests for trial particulars. It is possible, though somewhat doubtful, that the appellant hoped that the respondent would not find out that he beneficially owned shares in Blouvlei, Mont du Toit and Clos du Toit. Apart from what we have already said, there is the fact that the respondent obviously knew that the appellant was closely concerned with the operations of these companies. Given the acrimony which was present by the time divorce proceedings began, the appellant could not have believed that the respondent’s team would not fully investigate the shareholdings and find the documents she eventually obtained through subpoenas.
[86] But even if he was deliberately concealing these matters, the facts eventually came out. They do not justify the further step of finding that he is not the beneficial owner of only 26 per cent of Blouvlei but of 100 per cent. A more likely explanation for the appellant’s unsatisfactory discharge of his obligations in respect of discovery and trial particulars is a combination of arrogance, obduracy and a personal conviction that the respondent was not entitled to maintenance and that all his assets of substance were excluded assets. He saw the respondent as trying to draw him into another long and ugly divorce in circumstances where (as he saw things) she had no legitimate claim to maintenance or accrual. He thus dealt with her and her legal team in a high-handed and cavalier fashion. This is inexcusable but, as stated, does not warrant the summary rejection of his version as it bears on the question of the beneficial ownership of assets.
Are appellant’s interests in the companies excluded assets?
[87] The next question is whether the appellant’s shareholdings in Blouvlei (26 per cent), Mont du Toit (25.1 per cent) and Clos du Toit (26 per cent), and his loan accounts in these and other companies, are excluded assets.
[88] In regard to the shares, it is common cause that there is no equity in Mont du Toit and Clos du Toit. The value of the shares in Blouvlei are a function of the value of the Wellington farm. Both sides adduced expert evidence of such value. The trial court preferred the evidence of the respondent’s expert, providing reasons for its conclusion. We have not been persuaded that we should interfere with this finding. We thus accept the trial court’s finding that the Wellington farm is worth R15 689 185.
[89] It is also common cause that it would be appropriate, in valuing the shares in Blouvlei, to deduct the company’s indebtedness on loan account in the amount of R1.55 million and the capital gains tax which would be incurred to sell the farm so as to unlock the value in the shares. The net value of the shares after making these deductions is R10 924 786. The value of the appellant’s 26 per cent shareholding in Blouvlei is thus R2 840 444. The question is whether this asset is an excluded asset.
[90] The loan accounts in South African companies, which the respondent contends should be brought into account as part of the appellant’s accrual, are: (a) R466 806 in Caledon Street Guest House; (b) R858 952 in Mont du Toit; and (c) R1 550 000 in Blouvlei. These three amounts total R2 875 758. In relation to Namibian companies, the respondent seeks to bring into account a net amount of R1 245 028 in Gamsberg (after deducting an amount which appellant owes Schoongezicht).
[91] The treatment of the amount of R1.55 million as an amount owed on loan account to the appellant was based on a table prepared by Mr Greenbaum in his second report. However, the company’s financial statements record the loan account creditor as Gamsberg, not the defendant. Mr Greenbaum was alerted to this during cross-examination and he corrected himself, saying that it appeared the appellant had advanced the funds to Gamsberg which in turn lent the money to Blouvlei. This seems to be correct. Ms Ladopoulos did not show the appellant as having a loan account in Blouvlei. The appellant testified that he was not a creditor of Blouvlei and that he had advanced money to it through Gamsberg. This was not challenged. Mr Greenbaum’s schedule correctly reflects the appellant’s claim on loan account against Gamsberg in the amount of R1 514 032 (reduced to R1 245 028 after netting off an amount owed by the appellant to Schoongezicht). It appears to follow that Greenbaum’s inclusion of a loan claim by the appellant against Blouvlei in the amount of R1.55 million is incorrect and constitutes double-counting.
[92] It was for the appellant to prove on a balance of probability that the shares and loan accounts, all of which came into existence after the conclusion of the marriage, were excluded assets. Apart from his direct evidence that the shares and loan accounts were funded from the realisation of excluded assets, he relied on Mr Greenbaum’s analysis of such proceeds.
[93] That the appellant would have used excluded proceeds to fund the shares and loan accounts is not implausible. If his strategy was to minimise his non-excluded estate, it would have made sense for him to use the proceeds of excluded assets to fund the acquisition of other assets and to use his professional income (which was his main non-excluded resource) to pay for living expenses and operational farming expenses. Although it might have been mean-spirited, it would not have been unlawful for the appellant to have minimised his accrual estate in this way. Contemporaneous notes which he wrote during July 1997 and June 1998 are consistent with such a strategy.
[94] On the other hand, the appellant did not produce documents which provided a paper trail by which one could follow the realisation of particular excluded assets and the application of the proceeds in the acquisition of the farm and the advancing of funds to the companies. The bank statements and paid cheques, particularly for his No 2 account, were missing for the crucial years. The appellant claimed that he had been unable to find them despite diligent search. As against this, he was able in the main to produce isolated bank statements and cheques (or cheque stubs) to vouch for the property expenditure forming the subject of his counterclaim, despite the fact that such expenditure fell within the period for which in general he was unable to produce bank records.
[95] The appellant’s supposed inability to produce these documents is certainly a matter of surprise. He seems to have retained other documents which supported his counterclaims. The marriage was never so stable and happy that he would have discounted the possibility of a future divorce. His contemporaneous notes of July 1997 and June 1998 show that he was anxious to record matters relevant to accrual. Why then would he not have been careful to retain the bank statements for his No 2 account, particularly for the critical period when on his version the excluded assets were being realised to fund other acquisitions?
[96] His bank records are not the only relevant documents which the appellant was unable to produce. He did not discover complete sets of his personal financial statements or the financial statements of the companies over the relevant period. It is not possible to know exactly when, for example, the loan accounts came into existence and what their balances were from time to time. There is no information whatsoever as to the appellant’s professional income in South Africa for the tax years 1993 to 1996 or regarding his professional income in Botswana and Lesotho.
[97] In regard to Mr Greenbaum’s analysis, which sought to link the proceeds of realised assets to the acquisition of shares and loan accounts, he was in the nature of things dependent to a considerable extent on the reliability and completeness of the information he received from the appellant. Furthermore, his assumption – that substantially all the proceeds of the excluded assets must have been invested in the shares and loan accounts, and that the appellant’s professional income would not have sufficed – is to a considerable extent undermined by evidence which was not available to him when he prepared his reports.
[98] More particularly, it was put to him that on the evidence, including that of the appellant himself, substantial amounts from the proceeds of excluded assets were applied to known purposes other than the acquisition of the shares and loan accounts. According to the respondent’s analysis, the amounts applied to such other purposes totalled around R8.1 million, made up as follows:
(a) the donation of R300 000 to the respondent in fulfilment of the provisions in the contract;
(b) R289 000 used by the appellant from the No 2 account to ‘square’ the overdraft on his business account (ie reduce it to nil) at the commencement of the marriage;
(c) R136 877 from the No 2 account to finance renovations to Saxonwold;
(d) About R620 000 which the appellant ‘certainly paid (his) creditors’, in South Africa and Namibia;
(e) R720 000 from the sale of Schoonwater and/or Beerzijnbosch (the two farms in Mpumalanga) which were used to pay off the appellant’s Lloyds liability;
(f) R1.7 million (the entire value of Namibian Factors, a company in Namibia previously owned by the appellant), which was effectively donated to the family trust;
(g) R1.5 million (being the proceeds of the sale of the Muy Bien properties in Windhoek), which was effectively paid abroad;
(h) R42 000 in STC (company tax) paid from the proceeds of the Brenton-on-Sea apartment registered in the name of Emtilist (Pty) Ltd;
(i) R172 000, which the appellant paid for his children’s private school education in advance;
(j) R82 544 with which the appellant reimbursed his business account from the No 2 account, as per his note of July 1997;
(k) R630 000 transferred from his No 2 account to his personal account to reduce the overdraft on the latter account, as per his note of June 1998;
(l) R1.92 million paid by the appellant to register trade marks acquired by Clos du Toit.
[99] It is not entirely clear from the evidence that all of these amounts were paid from excluded proceeds. For example, the donation of R300 000 may have come from the appellant’s professional income. In regard to the trade marks, the evidence did not show that R1.92 million was actually spent. The appellant testified that the Clos du Toit loan account in this amount was based on a Capital Gains Tax valuation and that, while the costs of registering the trade marks were not insubstantial, they were considerably less than R1.92 million. It is also not clear that the expenditure on registering the marks was not funded from excluded proceeds. Furthermore, Mr Greenbaum took into the expenditure on the trade marks and the payment of the Lloyds liability from excluded proceeds. Even so, the excluded proceeds used for purposes other than the shares and loan accounts exceed R4.55 million.
[100] Having regard to the absence of a paper trail, question marks as to the probity of the appellant’s discovery and the trial court’s adverse credibility findings, we cannot find that the appellant discharged the onus of proving that the shares and loan accounts were in their entirety funded from excluded assets. We think it is likely that some part of the proceeds of excluded assets was so applied; it does seem unlikely that the appellant’s professional income would have been sufficient, particularly if such income was also being used to fund living expenses. However, once we find – as we do – that the appellant failed to prove that the shares and loan accounts in their entirety were funded from excluded assets, he provided no basis on which an apportionment could be made. That would be entirely a matter of speculation.
[101] Accordingly, the shares and loan accounts, in the values we mentioned earlier, should be included in the appellant’s accrual. If we had found that the shares and loan accounts were to any extent excluded assets, it would have been necessary to consider, in the light of the concluding part of clause 5 of the contract, what adjustments should be made by virtue of the fact that the appellant used his professional income to improve the Wellington farm by the planting of vineyards in 2008 and 2013. The need to undertake this enquiry falls away in the light of our conclusion that the shares and loan accounts are in any event not excluded assets.
The living annuity
[102] As stated, the high court included a Sanlam Glacier living annuity as part of the appellant’s assets for purposes of calculating the accrual. This annuity is payable in terms of an annuity contract between the appellant and Sanlam (the Glacier contract). At the time of the high court’s judgment the Glacier contract was supposedly worth R3 270 368. The high court did not furnish any reasons for this decision – it simply included the annuity in its calculation.
[103] The facts with regard to the Glacier contract are as follows. The appellant had a number of retirement annuities which dated back to the 1960s. When he turned 60 (ie in March 1999), he elected to apply his maturity value in the retirement annuities to the purchase of a living annuity in his own name from Sanlam to provide himself with a monthly income which was intended to serve as his pension.
[104] The main question in M v M[26] was whether a living annuity forms part of the estate of the annuitant spouse for the purpose of assessing accrual. Significantly, the parties had reached agreement in that case that a living annuity is not a pension interest as defined in the Divorce Act[27]. Victor J rejected the contention that the living annuity fell within the annuitant’s estate. The living annuity (coincidentally also a Sanlam Glacier product) was purchased during the course of the marriage and was used by the annuitant as a monthly source of income. As was stated, the question whether a living annuity on divorce forms part of the accrual is novel. M v M is to our knowledge the only decision thus far where the question has been dealt with in a fully reasoned judgment.
[105] Counsel for the respondent relied on a number of cases where it was held that an accrued pension is an asset of the joint estate of parties married in community of property.[28] We are not called upon to decide the rights of spouses in connection with pension interests acquired by one of the spouses during the course of a community marriage. None of the cited cases concerned a living annuity or the question whether the underlying capital value was an asset belonging to a spouse for purposes of an accrual calculation. They are therefore no authority for the contentions advanced by the respondent. It was contended further on behalf of the respondent that M v M did not refer to any of these cases and was wrongly decided.
[106] In order to qualify as a ‘living annuity’ for income tax purposes, an annuity contract must comply with certain requirements. These are currently contained in the definition of ‘living annuity’ in s 1 of the Income Tax Act 58 of 1962 read with Government Notice 290 of 11 March 2009[29]. Although the Glacier contract is not part of the record, it is not in dispute that the appellant’s Glacier contract meets these requirements. The value of the annuity is determined solely by reference to the value of the assets specified in the Glacier contract (the opening value would have been the amount applied to purchase the annuity in March 1999 less initial costs). In other words, the amount of the annuity is not guaranteed. The assets themselves belong to Sanlam, fluctuate with market conditions and are reduced as the annuity is drawn down. The annual amount which the appellant can draw as an annuity is not less than 2.5 per cent and not more than 17.5 per cent of the current capital value. On the appellant’s death, Sanlam must pay any remaining capital to the appellant’s nominee as an annuity or lump sum. If there is no such nomination, the capital must be paid as a lump sum to the appellant’s estate.
[107] The Glacier contract does not result in the appellant being a member of a ‘pension fund organisation’ as defined in the Pension Funds Act 24 of 1956. His status as such terminated when his interests in his previous retirement annuity funds were applied to purchase the living annuity. The provisions in the Divorce Act dealing with a spouse’s ‘pension interest’ are thus not applicable. It appears to be generally accepted in the pension fund industry that the provisions of ss 37A to 37D of the Pension Funds Act apply to a living annuity purchased in the name of a former member of a retirement annuity fund (see Kobus Hanekom Manual on Retirement Funds and Other Employee Benefits 2015 para 19.1.10 at 932). Both forensic accountants who testified in the present case gave this as their understanding. The appellant’s counsel took this for granted in their heads of argument, and the contrary was not submitted in the respondent’s counsel’s heads of argument or in their supplementary note dealing with accrued pension claims.
[108] Having regard to the nature of the Glacier contract, we are of the view that its supposed capital value cannot be included as part of the appellant’s accrual. The capital belongs to Sanlam, not the appellant. The appellant’s only contractual right is to be paid an annuity in an amount selected by him within the permissible range specified by law. His right to receive any particular annuity instalment is subject to a condition of survivorship, ie that he should be alive on the date on which the next annuity instalment becomes payable. If he does not survive to the next date, the fate of the capital will be determined by whether or not he has nominated a beneficiary. The capital may or may not be paid to his estate, depending on whether or not there is such a nomination.
[109] If the supposed capital value of the Glacier contract were included in the appellant’s accrual, one would have the anomalous outcome that he would be obliged to pay half its value to the respondent in circumstances where he has no right to claim half the capital from Sanlam. He would have to satisfy this part of an accrual award from other assets. While in the present case the appellant may have other assets from which to make payment, the question is one of principle. If the Glacier contract is to be included in the appellant’s accrual, it would have to be included in the accrual of any spouse with a comparable annuity contract, even though the contract were such spouse’s only ‘asset’. The outcome would be even more anomalous if the spouse’s interest in the annuity contract was exempt from attachment in terms of s 37A of the Pension Funds Act, because then there would be nothing for the other spouse to attach in satisfaction of the accrual award.
[110] It might be argued that the appellant’s conditional right to future annuity payments is an asset which can be valued. However, and even if this were an asset which did not enjoy the protection of s 37A and could notionally be included in a spouse’s accrual (on which it is unnecessary to express an opinion), the respondent did not adduce evidence to establish the value of the conditional right. Such value could not simply be equated to the value of the capital held by Sanlam. The value of the conditional right would be affected by the appellant’s life expectancy and the rate at which he has in the past, and is likely in the future, to draw down his annuity.
[111] Indeed, the respondent did not even prove the capital value held by Sanlam at the relevant date. She relied on a Sanlam letter which gave the fund value used for illustrative purposes as at 24 July 2013. Assuming that that was the actual capital held by Sanlam as at that date, this was more than three years before the trial court gave judgment. The appellant’s estate had to be valued as at the date of the divorce. In the intervening three-year period the capital value would have reduced by the amount of annuity instalments paid to the appellant and would have fluctuated (either up or down) by virtue of market conditions
[112] It follows that the high court erred by including the Glacier contract as part of the appellant’s accrual. The monthly income derived by the appellant from the annuity, however, forms part of his total income which has a bearing on his means to pay maintenance, if any, to the respondent.
Miscellaneous aspects of accrual
[113] The accrual can only be calculated once the counterclaims and the effect of the restitution of donations have been determined. With regard to the rest of the appellant’s estate for accrual purposes, the list of assets is mostly uncontentious. We will deal with the Rondebosch house presently. It is common cause that the appellant owns two immovable properties in Wellington, namely a house in a security estate valued at R500 000 net (R2.2 million less a bond of R1.7 million) and a vacant erf valued at R100 000. We have already dealt with the related-party loan claims. The rest of the assets comprise personal effects, policies (excluding the living annuity) and net investments in the appellant’s legal practice and farming operations. We have excluded the contents of safety deposit boxes in Paarl and Hamburg, since we are not satisfied that the contents of these boxes all belong to the appellant. There is insufficient evidence, in particular with regard to the Hamburg box, of the appellant’s ownership. From the calculation must be deducted the legal fees owing to the appellant’s attorneys, which the high court assessed at R2 million.
[114] We deal next with the donations under the contract and the appellant’s counterclaims, since they may have an effect on the final calculation of the accrual. Before dealing with the appellant’s counterclaims, it is necessary first to consider the effect of the declaration of invalidity of the waiver clause on the reciprocal donations made by the appellant.
Restoring the donations under the contract
The cash donation of R300,000
[115] The declaration of invalidity of the waiver must self-evidently result in the restoration of the donations. In principle, the appellant is entitled to recover the donations he made in return for the waiver. He claimed restitution of the donations as an enrichment claim by way of the condictio indebiti. The legal basis for that claim is that the waiver provisions in the contract were void ab initio. That must be so, given the majority’s finding that the clause is against public (legal) policy. On Rogers AJA’s approach as to why the waiver should be overridden in this case, the preferred approach might be to apply the rules of restitutio in integrum.[30] The respondent, having elected not to be bound by the waiver and having persuaded the court to exercise its overriding discretion, must restore what she received as quid pro quo for the waiver. Ultimately, regardless of which one of these two remedies one prefers, the outcome will be the same.
[116] The respondent tendered to return the donations in the event of the waiver being declared invalid. The high court correctly found that the appellant had in fact made the cash donation of R300 000. Based on the condictio, the respondent is presumed to have been enriched and, absent a plea of non-enrichment or loss of enrichment as a defence, the respondent must restore that amount.[31] An actuarial calculation shows that the present day value of that amount is R1 409 596 – that is the amount which should be restored, so it was contended. But there was no evidence that the respondent invested the money in a manner that would have generated a return equal to the inflation rate. Interest on a claim based on unjustified enrichment is not payable until the debtor has been placed in mora.[32] Since the appellant’s counterclaim was conditional upon the respondent succeeding in having the waiver declared invalid, the appellant is in our view entitled only to mora interest on the amount of R300 000 from the date of the high court’s judgment.
The donation of a half-share in the matrimonial home
[117] The high court ordered the appellant to transfer to the respondent his half share in the Rondebosch property against payment of the sum of R2 million by the respondent, to be made by way of a reduction of the accrual award. The parties are registered as co-owners of this property. The appellant claimed termination of the co-ownership under the actio communi dividundo. A determination must, however, first be made as to the true status of the Rondebosch property, having regard to the fact that it is the third of the parties’ successive matrimonial homes, the first of which was the subject of a donation in the contract.
[118] It will be recalled that one of the donations required to be made in terms of the contract was a half share in the Twickenham property, owned by the appellant at the time of the marriage. The trial judge upheld the contention advanced on behalf of the respondent that this donation did not take place because the half-share was never registered in her name. We disagree. In terms of the contract, once the contract had been executed and registered, the donation itself was completed as between the parties. What was required further in terms of clause 6.1 was for the appellant, upon a request by the respondent, to transfer the property into their joint names. Such a request for transfer was never made. The parties never intended to live in the Twickenham property. Instead, by agreement, the respondent began looking for another house.
[119] The Twickenham property was sold in May 1993 and the parties agreed to apply the proceeds of that sale to buy the Saxonwold property. The respondent was registered as a co-owner of that property and she must undoubtedly have regarded that as fulfilment of the promised donation in the contract. In her evidence she indicated that perhaps for ‘practical reasons’ she had not been registered as a co-owner of the Twickenham property. And she acknowledged that ‘he used the money from the first house as my share to buy the next house’. This evidence accords with the respondent’s plea to the appellant’s conditional counterclaim, where it was admitted that her half-share of the proceeds from the Twickenham property was applied to buy the Saxonwold house. She never testified that the donation in the contract had not been fulfilled.
[120] The proceeds of the sale of the Saxonwold property in April 2007 were by joint decision applied to buy the Rondebosch house. The respondent was again registered as a co-owner. It is common cause that she contributed no funds of her own to the purchase or improvement of the Saxonwold or Rondebosch properties. Applying the principles of unjustified enrichment or restitutio in integrum, self-evidently the respondent must restore her half-share of the Rondebosch property, which is directly traceable to the donation of a half-share of the Twickenham property. She has been enriched, by that donation, in the amount currently represented by her half-share of the Rondebosch property. That amount equals the extent of the impoverishment of the appellant, because on each succeeding occasion the matrimonial homes were registered in the names of both parties where, in truth and in fact, they should have been registered in the appellant’s name alone, given the invalidity of the donation. (The Parkview apartment does not feature in this discussion, because the appellant bought the respondent’s half-share in April 2014 for the sum of R320 000.)
[121] The high court’s order entitling the respondent to buy the respondent’s half share of the Rondebosch property thus cannot stand. Conversely, though, the appellant’s accrual must be determined on the basis that he has the sole right to the Rondebosch property. This raises the question as to whether and to what extent the Rondebosch property is an excluded asset.
[122] The Twickenham property was an excluded asset. That is not in dispute. The proceeds from its sale, which were also excluded proceeds, were applied to purchase the Saxonwold property. The appellant also used some excluded resources to make improvements to the Saxonwold property. On the other hand, he used his professional income (a) to make other improvements to the Saxonwold property; (b) to make mortgage payments on the bond taken out to fund the balance of the Saxonwold purchase price and to fund some of the improvements; and (c) to pay rates and taxes on the property. When the Rondebosch property was bought from the Saxonwold proceeds, he used his professional income to pay for certain property-related expenditure. The value of the Rondebosch property is thus partly referable to excluded resources and partly referable to non-excluded resources.
[123] It is thus clear that at least part of the Rondebosch property’s value should be regarded as an excluded asset. How is this to be determined? The interpretation of the concluding part of clause 5, which we quoted earlier, is not free from difficulty. If non-excluded funds are applied in the manner contemplated, must they be included in the accrual at the nominal amount expended or with reference to the amount by which the expenditure has led to an enhanced value at the date of the divorce? The clause also does not explicitly deal with the case where an asset, from the outset, is purchased partly with excluded resources and partly with non-excluded resources. If one spouse were to purchase an asset for a price of R1 million and were to pay the price as to R500 000 from excluded resources and as to R500 000 from non-excluded resources, and if at the date of the divorce the asset were worth R5 million, would one include R500 000 or R2.5 million in the accrual? The same question would arise if, after the marriage, one spouse were to spend R500 000 from non-excluded resources in improving an existing excluded asset then worth R500 000.
[124] Although the clause says that the ‘funds’ shall be included in the calculation of the accrual, it does not specify how they must be factored into the calculation. We do not think the only meaning of which the clause is linguistically capable is that the funds must be included at their nominal amount at the time of the expenditure. Such a meaning strikes us as unjust and unbusinesslike. Why ascribe all the enhancement in value to the excluded component? We thus think that where both excluded and non-excluded resources are applied to an asset, the extent to which the value of the asset at the date of the divorce is to be included in the accrual must be determined by assessing a fair and reasonable ratio between the excluded and non-excluded resources which contributed to the asset’s value. While in many cases this would result in the inclusion of a greater amount than the nominal non-excluded expenditure, the converse would also be possible. If the asset decreased in value, the amount to be included in the accrual might be less than the nominal non-excluded expenditure.
[125] As to the onus of proof, the appellant, as we have said, bore the onus of proof to establish exclusion. As to the concluding part of clause 5 (which qualifies the extent of the exclusion), there may be an evidentiary onus on the non-owning spouse to show that there was non-excluded expenditure on the excluded asset. Once, however, it appears from the evidence that only part of the asset qualifies for exclusion, considerations of fairness and pragmatism dictate that the owning spouse bears the onus of proving the extent of the exclusion.
[126] The determination of the excluded and non-excluded components must inevitably be a somewhat rough and ready process. The Twickenham property yielded net proceeds of R372 000 in May 1993. The Saxonwold property was purchased at the same time for R470 000. Subsequently a bond of R250 000 in favour of Nedbank was registered over the Saxonwold property. There is some confusion in the evidence as to what occurred in the period after the Saxonwold property was transferred to the parties but before the registration of the mortgage bond. It seems that the appellant was granted a facility by the bank which was then substituted by the mortgage bond. We thus proceed on the basis that R250 000 was borrowed to cover the balance of the purchase price of the Saxonwold property and some of the improvements. In May 2001 a mortgage bond in favour of Investec was registered in place of the Nedbank bond. At that time the balance owing on the Nedbank bond was R245 156.
[127] In his counterclaim for property-related expenses (as adjusted during the course of the trial), the appellant alleged that he spent R490 528 on Saxonwold improvements; made bond payments totalling R541 513 up to the time the Saxonwold property was transferred in August 2007; and paid rates and taxes totalling R143 103 over the period August 1993 to May 2007:
(a) In regard to the improvements of R490 528, the appellant testified that R136 877 came from his No 2 account, ie from excluded resources. He knew this because he kept a list at the time which was handed in as evidence. These payments were made during 1993. Regarding the balance of R353 651, he was unable to say from which accounts they were paid. Given the incidence of onus, we must assume in favour of the respondent that the balance was paid from non-excluded resources.
(b) The bond payments were not made from excluded resources. The appellant testified that the Nedbank bond payments were made from his private account and the Investec bond payments from his practice account.
(c) The appellant testified that the rates and taxes were definitely not paid out of the No 2 account, ie from excluded resources. In our view, rates payable on property constitute a ‘debt’ which is ‘related to’ the property within the meaning of the concluding part of clause 5. Although the rates and taxes would have included a component for services such as water and refuse collection (which are more in the nature of living expenses than debts relating to the ownership of the property), the evidence does not enable one to isolate these components. Accordingly the full expenditure on rates and taxes must be included.
(d) The nominal amount of the excluded expenditure on the Saxonwold property was thus R508 877 (the Twickenham proceeds of R372 000 plus Saxonwold improvements of R136 877) while the nominal amount of the non-excluded expenditure was R1 038 267.
[128] When the Saxonwold property was sold, it yielded net proceeds of R5 577 477 after settlement of the Investec bond liability and agent’s commission. In order to determine what portion of the net proceeds of R5 577 477 constituted an excluded asset, it is necessary to ascertain what proportion of that amount is fairly referable to excluded and non-excluded resources respectively. In the absence of any better method, we think it would be fair to determine this proportion by adjusting the amounts in question to account for inflation. Adopting this approach, the excluded and non-excluded components of the net Saxonwold proceeds amounted to 45 per cent and 55 per cent respectively. (So as not to burden this judgment, the workings are set out in appendix 1.)
[129] Out of the net proceeds of R5 577 477, an amount of R5 231 664 was used to fund the purchase and to improve the Rondebosch property while the balance was expended on the Parkview apartment (which was funded mainly by a mortgage bond). It is reasonable to assume that the excluded and non-excluded components of the Saxonwold proceeds were applied in the same proportions to the Rondebosch and Parkview properties. The appellant paid rates and taxes on the Rondebosch property totalling R140 103. Since it was not shown that these payments came from non-excluded resources, they must be included in the calculation of the accrual. This results in a final ratio, applicable to the Rondebosch property, of 44 per cent excluded and 56 per cent non-excluded. (Again, so as not to burden the judgment, the workings are contained in appendix 1.)
[130] At the start of the trial in August 2014 the parties agreed that the Rondebosch property was worth R5.6 million. It follows that we treat 56 per cent of that amount, namely R3.136 million, as part of the appellant’s accrual. There is a mortgage bond registered over the Rondebosch property and for purposes of calculating the accrual the parties accepted the liability as being R800 000. The question arises as to whether the whole of this liability, or only 56 per cent thereof, is deductible in arriving at the appellant’s net accrual.
[131] The MPA is silent on the point. If a person has incurred a liability to acquire or improve an excluded asset, we would think the liability, like the asset, should be left out of account. In the present case, however, the liability secured by the mortgage bond was not incurred to acquire or improve the Rondebosch property. It was incurred later, when the appellant urgently needed money to pay tax. Although the loan is secured by a mortgage bond, the tax liability which caused the appellant to borrow the money is associated with his professional income, which is not an excluded resource. We thus think the full liability is deductible (the contrary was not argued). For the same reason, the payments the appellant has made to the bank in respect of the liability secured by the Rondebosch mortgage bond should not be taken into account for purposes of the last part of clause 5 of the contract.
The counterclaim for property-related expenditure
[132] The appellant claimed the repayment of half of all the amounts spent in respect of jointly owned properties. The claim was based on co-owners’ joint responsibility for costs related to jointly owned property. However, in view of our finding that the donations in the contract must be unwound, the counterclaim for property-related expenditure falls away, since in effect the money was expended on property which belonged to him exclusively. In supplementary heads it was conceded on behalf of the appellant that, if the half share of the Rondebosch property is restored to the appellant, his counterclaim for property-related expenses falls away.
The counterclaim for delivery of movables
[133] The appellant claimed the delivery of certain movables which had remained in the Rondebosch house. The respondent disputed that all items were in her possession but has, nonetheless, tendered the return of all the movables on the list which are still in the house, once the accrual award due to her has been paid.
Final calculation of accrual
[134] Based on the findings made thus far, the appellant’s net estate for purposes of computing the accrual is as set out in the following table:
Included assets |
|
Included component of Rondebosch property |
3 136 000 |
26% shareholding in Blouvlei |
2 840 444 |
SA loan accounts[33] |
1 325 758 |
Net Namibian loan accounts[34] |
1 245 028 |
Erf 11367 Wellington (net of bond liability) |
500 000 |
16 Albedo Street, Wellington |
100 000 |
Legal practice assets |
959 137 |
Farming assets |
500 000 |
Personal effects |
1 095 000 |
Debtors |
3 947 |
Policies |
156 102 |
Total accrual assets |
11 861 416 |
Liabilities |
|
Rondebosch bond |
(800 000) |
Overdraft |
(168 934) |
Legal costs |
(2 000 000) |
Total liabilities |
(2 968 934) |
Net accrual |
8 892 482 |
Plaintiff’s half-share |
4 446 241 |
[135] As against the respondent’s accrual entitlement of R4 446 241 (which would carry mora interest from date of judgment), the respondent became indebted to the appellant in the amount of R300 000 plus mora interest from date of judgment. The appellant admitted, however, that he owed the respondent €23 000 for monies she lent him in respect of the French farm. Although he sought to set off this indebtedness (in a rand amount of R230 000) against his counterclaim for improvements, on the approach set out above he has no counterclaim for improvements. We would thus set off his indebtedness of R230 000 against the respondent’s obligation to repay the donation of R300 000, giving a net amount of R70 000 owed by her as monetary restitution. Her liability of R70 000 should, in turn, be set off against the appellant’s accrual liability to her, resulting in a net accrual liability owed by the appellant to the respondent of R4 376 241 together with mora interest from date of judgment.
[136] The respondent will also have to co-operate in having her half share of the Rondebosch property transferred to the appellant. We do not know whether transfer duty will be payable. Certainly some legal costs will be incurred. The parties should share these costs. This can be achieved by authorising the appellant to deduct, from his net accrual obligation of R4 376 241, a half share of the reasonable costs of securing transfer of the respondent’s half share of the Rondebosch property into his name.
The respondent’s maintenance claim
[137] On 8 December 2016 Baartman J made an order that, pending this appeal, the appellant pays the sum of R7 800 per month as maintenance to the respondent as well as the bond and municipal charges in respect of the Rondebosch property. That order remains extant and, we were informed from the bar, is being complied with.
[138] An assessment of the respondent’s claim for maintenance must take into account the factors mentioned in s 7(2) of the Divorce Act 70 of 1979. She claimed certain amounts of maintenance unconditionally and certain other amounts conditionally upon her being awarded less than R4.4 million by way of accrual. Her unconditional claims, which for convenience we shall call Part A, were in summary the following:
(a) that the appellant pay her R30 000 per month, increasing annually by the percentage change in the ‘Headline inflation rate (also known as the Headline Consumer Price Index) as notified by Statistics SA’ (the index);
(b) that the appellant bear the cost of all her medical and health-related treatment;
(c) that the appellant provide her with a new vehicle of her choice, with a value equivalent to a new Subaru Outback 2.5i Premium, every five years.
Her conditional claims, which for convenience we shall call Part B, were in summary the following:
(a) that the appellant purchase a home of her choice for the maximum price of R3.8 million, escalating at 15 per cent per annum as from 1 August 2011 until the property is made available to her;
(b) that the appellant pay all costs of transfer into her name and furniture removal costs;
(c) that the appellant pay all rates and taxes on the property, houseowner’s insurance premiums and the costs of interior and exterior maintenance;
(d) that the appellant pay her R250 000 to enable her to furnish and equip the new home.
[139] In respect of the Part A claims, the high court awarded the respondent R30 000 per month increasing annually in accordance with the index and a new vehicle with the claimed value every six years. The court made no order on the claim in respect of medical expenses. Because the high court awarded her more than R4.4 million by way of accrual, no order was made on Part B. Since the respondent’s accrual award in the present case (prior to setting off her net restitutory obligation of R70 000) is R4 446 241, we likewise need not consider the Part B claims.
[140] The respondent was 51 years old at the time the high court gave judgment and is now 53. The appellant was 77 years old and will be 79 by the time judgment in the appeal is delivered. The appellant’s evidence is that, but for his trial expenses, he would by now have retired. It is not reasonable to expect him to continue working for any length of time at his age. The high court misdirected itself by finding that there was no reason why his real net taxable income in the future would not be as high as it was over the period 2012 to 2015.
[141] The appellant’s ability to pay maintenance must thus be judged on the assumption that he will not continue to earn professional income. He will continue to receive income from his living annuity but this is unlikely to cover more than his own living expenses. His farming operations may become profitable in the future and Mont du Toit’s wine-making operations may also become profitable and result in the payment of dividends to the shareholders. However, neither the farming nor the wine-making operations are yet profitable. It is thus to the appellant’s capital resources that regard must mainly be had. For maintenance purposes, these include both excluded and non-excluded assets.
[142] As to the respondent’s reasonable living expenses, her itemised maintenance requirements came to R37 113 per month. She arrived at the claimed amount of R30 000 by deducting from R37 113 the items for rates and taxes, houseowner’s premiums and medical expenses. The first two deductions were made on the assumption that, if she were awarded less than R4.4 million by way of accrual, the appellant would be responsible for paying rates and taxes and houseowner’s premiums; and that if she were awarded more than R4.4 million she would not reasonably require the appellant to pay these expenses. The third deduction was made on the basis that she advanced a separate claim for medical costs.
[143] The high court’s acceptance of the respondent’s claim of R30 000 per month was based on a finding that it was reasonable for the respondent to continue living in the Rondebosch house. This was a misdirection. The Rondebosch house is a family home which was bought at a time when the parties were still married and the children were living at home. By the time of the trial both children were adults and had left the home. The appellant has always paid their reasonable needs. The appellant has downscaled and the respondent can reasonably be expected to do likewise. There was evidence that the respondent earns money by accommodating a boarder attending a nearby school. This is not something she is entitled to do at the appellant’s cost.
[144] A townhouse or flat in the southern suburbs of Cape Town with a main bedroom and a guest room would be adequate for her reasonable requirements. On this basis, some of the respondent’s maintenance claims must be moderated. With reference to her itemised schedule totalling R37 113, the rates and taxes and houseowner’s premiums must be deducted, given the accrual award. We would reduce the amount claimed in respect of electricity and property repairs/repainting by one-third. Her grocery claim of R7287 includes groceries for children, including presumably the boarder. Since the children no longer live at the home and the appellant cannot be expected to maintain boarders, we would reduce the sum to R5000. The monthly claim of R400 as pocket money for the children should also be disallowed. This reduces the total claim from R37 113 to R28 514 which we will round down to R28 500.
[145] In regard to medical expenses, we consider it preferable to include an allowance for medical expenses in the monthly maintenance rather than to lumber the appellant with an open-ended liability such as contemplated by the respondent’s separate claim in that regard. Her schedule of monthly expenses reflected medical costs totalling R3385 per month, the major component being a monthly medical aid contribution to Discovery of R3028. Although her current employer pays half of this amount, it is deducted from her salary. Since the salary figures canvassed by the experts were, we assume, prior to such deductions, we shall not make any downward adjustment in respect of her monthly medical claim.
[146] Evidence was led regarding the respondent’s future earning capacity. After the separation in 2010, the respondent worked for some time for a Swiss wine company, known as 4G. She claimed to have done so for no remuneration. At the time of the trial the respondent was working at the online retailer, Amazon’s, call centre, rendering service to German-speaking clients who called in. That job entailed working around 30 hours per week for remuneration of between R6 000 and R8 000 per month. On behalf of the respondent, an industrial psychologist, Ms Elizabeth Hofmeyr, and a psychiatrist, Dr Konrad Czech, were called. Ms Hofmeyr opined that, once the respondent had overcome her emotional vulnerability and mental instability after a year or so, she could pursue employment opportunities in the export, wine, marketing or hospitality industry where she could earn between R120 000 and R180 000 per annum. Thereafter, according to Ms Hofmeyr, in approximately five to six years’ time, she could earn approximately R300 000 per annum on a more senior level. Ms Hofmeyr’s report was dated 29 January 2014.
[147] Dr Czech had been treating the respondent for severe major depression since June 2012. His report was dated 31 October 2014. When he saw her on 27 August 2014, the respondent presented with symptoms of major depression and acute stress reaction which was in remission. According to Dr Czech, depression can result in cognitive impairment. In his opinion, the respondent would only be able to work after a two year recuperation period, and then it must not be stressful work.
[148] Mr Matthias Le Roux, a human resource practitioner, based in Paarl in the Western Cape and who specializes in the wine industry, furnished a report and testified in the appellant’s case. Like Ms Hofmeyr, Mr Le Roux conducted an interview with the respondent. He evaluated her post as a wine marketer on post level 8. Mr Le Roux regarded the respondent as employable in international wine marketing, capable of earning between R30 000 and R50 000 per month.
[149] Mr Le Roux’s evidence was subject to trenchant criticism by the respondent’s counsel. In turn, the appellant’s counsel was critical of Dr Czech’s testimony and report. The high court rejected Mr Le Roux’s evidence on the basis that ‘he clearly aligned himself with the [appellant] and I do not consider his evidence particularly helpful or objective’. The opinion that the respondent had an earning capacity of between R30 000 and R40 000 per month was ‘far-fetched and absurd,’ according to the learned trial Judge. He considered Dr Czech and Ms Hofmeyr to be unduly optimistic about the respondent’s ability to recover and to compete effectively in the job market. The learned trial judge held that the respondent had worked as a marketer for the Mount du Toit cellar wines ‘under conditions of sheltered employment’.
[150] Unlike the appellant, the respondent will be able to earn an income for some years to come. Even if she were only to work until the age of 65, she still had 14 years (as at the date of the high court’s judgment) to earn remuneration. At the time of the trial, as noted, the respondent was doing part-time work in the Amazon call centre earning around R7000 per month. This is the level at which the high court pitched her earning ability. That determination was lower than the assessment of both sides’ experts and lower than the respondent’s own assessment.
[151] The high court’s determination is not reasonably supported by the evidence. The respondent has a German legal qualification. German is her native tongue and she is fluent in English and reasonably proficient in Afrikaans. She has also undergone training as a mediator. She has an engaging personality. In the latter years of the marriage she obtained knowledge about wine and was successful in expanding the international market for Mont du Toit’s wines. Her CV described her ten years with Mont du Toit as ‘international marketing’ in which she established markets in the United States and various European countries. When she travelled overseas for Mont du Toit, her hourly rate translated into monthly remuneration of around R30 000. To describe her work for Mont du Toit as ‘sheltered employment’, as the trial judge did, is demeaning and unwarranted.
[152] Although she is probably no longer able to turn her legal qualification to account in professional practice, the fact that she obtained a German law degree would mark her out as a person with above-average intellectual abilities. Coupled with the other attributes we have mentioned, she ought not to experience difficulty in obtaining employment in the Western Cape in wine-marketing or the hospitality industry. Since the children are adults, the respondent will have the freedom to travel overseas if her work takes her there, just as she did when marketing Mont du Toit’s wines.
[153] The respondent may have confined herself to the call centre job because of the stress of the trial and the need to take time off to attend to matters concerning the trial. Once the divorce was granted, however, there was no justification for confining her earning capacity to R7000 per month. That was the amount she earned for part-time work. Even if she stayed with Amazon, she testified that she could apply for higher positions if she worked full-time. Based on the evidence, including the expert opinions of Ms Hofmeyer and Mr le Roux, we think that within a fairly short period of time she could command R20 000 per month (and we regard this estimate as conservative).
[154] We would thus assume earnings of R10 000 per month for the first year following the divorce, R15 000 per month for the second year and R20 000 per month thereafter. Deducting these amounts from her reasonable maintenance needs of R28 500 per month, we arrive at a reasonable maintenance claim of R18 500 per month for the first year following the divorce (from 1 September 2016), R13 500 per month for the second year (from 1 September 2017) and R8500 per month thereafter (from 1 September 2018) until death or remarriage.
[155] Since the respondent’s maintenance claims were clearly formulated on the basis that she would not require assistance with accommodation if she obtained an accrual award of R4.4 million or more, it is unnecessary to consider providing her with a further capital sum to assist her in obtaining accommodation. We simply observe that in our view a sum of R4.4 million will enable the respondent to purchase and equip a downscaled property in the southern suburbs. Until the Rondebosch property is sold and she receives her accrual award, the respondent’s accommodation needs can be met by requiring the appellant to allow her to continue living there rent-free.
[156] The fact that the respondent reasonably requires maintenance in the amounts set out above does not automatically lead to the conclusion that the appellant should be ordered to pay them. An important question is naturally whether the appellant can afford to pay maintenance in these amounts. Since he is entitled to sole ownership of the Rondebosch property, and since he does not require that property for his own residential needs, the property could be sold. Although the parties agreed a valuation of R5.6 million, the evidence established that the property was almost certainly worth more than R6 million by the time of the trial. Rondebosch real estate has increased in value since August 2016. Assuming that the appellant were able to sell the Rondebosch property for a net amount of R5.2 million after settling the mortgage liability, he would be able to pay the respondent her accrual award of around R4.4 million and be left with R800 000.
[157] Based on the respondent’s estimated life expectancy and an actuarial calculation, it was agreed in the high court that the capital sum required to fund maintenance for the respondent of R1000 per month for life was R285 000. According to the respondent’s supplementary submission filed after the hearing of the appeal, the required amount is now R234 100. On the latest figures, it follows that, in order to fund monthly maintenance for the respondent of R8500 as from 1September 2018 for the rest of her expected life, the appellant will require a capital sum of just under R2 million. As indicated above, R800 000 of this amount would be available from the sale of the Rondebosch property. (We leave out of account the maintenance payable over the first two years following the divorce since most of this period lies in the past and the appellant has – presumably from his professional earnings – paid a substantial part of that maintenance by virtue of Baartman J’s order.)
[158] According to the accrual calculations, the appellant has net recoverable loan claims against Caledon Street Guest House, Mont du Toit and Gamsberg totalling R2 570 786. If these loan accounts were repaid to the extent of R1.2 million, his net cash resources following the sale of the Rondebosch property would suffice to fund his maintenance obligations as from 1 September 2018.
[159] It is true that, as matters currently stand, the companies in which the appellant has loan accounts do not have liquid resources from which to repay him. However, and having regard to the value of Wellington farm, the loan accounts could readily be repaid if the farm were sold. This would also yield the appellant a liquidation dividend on his 26 per cent shareholding in Blouvlei. The Wellington farming operation is one about which the appellant seems to be passionate. We do not share the high court’s scepticism in that regard. It may seem hard on him to require the farm to be sold so that he can pay the respondent maintenance. On the other hand, he is advanced in years and it does not seem likely that he will be able to continue as an active farmer for much longer. There is no evidence that any of his children have displayed an interest in taking over the farming operation. The sale of the farm would thus not be a great injustice to the appellant.
[160] Since the appellant’s preference shares in Gamsberg give him voting control of that company, he effectively controls more than 75 per cent of the votes in Blouvlei, Clos du Toit and Mont du Toit. It thus lies within his power to cause those companies to cease their operations and sell the farms. Even if that were not strictly the case, there is no reason to suppose that Tartan and the Monaco trust would stand in the way of the sale if the appellant so wished. There is no other person associated with Tartan or the Monaco trust who would be able to take over the active management of the farms. In any event, if the appellant were to demand repayment of his loan accounts, the companies in question would have no choice but to sell the farm or face liquidation. Of course, it is possible that Tartan and the Monaco trust have other resources available to them of which we have no knowledge. If they and the appellant wish to continue the farming operations despite the need to repay the appellant his loan accounts, Tartan or the Monaco trust might be able to advance the relevant companies the necessary funds from such other resources.
[161] The marriage between the parties lasted 24 years. Because of the marriage, the respondent uprooted herself from Germany and was not able to pursue her legal career there. During a large part of the marriage the respondent was responsible for the primary care of the children because the appellant was pursuing his legal career, a career which would sometimes took him away from home. We do not think she can be criticised for having declined to requalify herself as a South African lawyer. That would have been a substantial undertaking as a middle-aged woman, given the significant differences between South African law and German law. She testified, and we have no reason to doubt, that she has now made her life in South Africa, more particularly in the Cape Peninsula, and does not wish to live in Germany.
[162] In all the circumstances, and having regard to the factors in s 7(2), we would make an order for monthly maintenance in the amounts indicated above. We would also order that the respondent be entitled to retain as her own the vehicle she was driving at the time of the high court’s judgment. We would not order the appellant to pay for the periodic replacement of the vehicle.
[163] The amounts of monthly maintenance mentioned in this judgment are stated in nominal terms as at 1 September 2016. They must be adjusted in accordance with the percentage change in the index in order to arrive at the actual amounts payable as from 1 September 2017 and 1 September 2018 respectively. Furthermore, the adjusted amount which is payable as from 1 September 2018 must be annually adjusted on 1 September of each succeeding year with reference to the same index.
Costs and order
[164] In regard to costs in the high court, the substituted order we intend to make would still represent substantial success for the respondent (as plaintiff), even though it is substantially less than the trial court awarded her. She thus remains entitled to her costs in the high court.
[165] The high court ordered the appellant to pay, on an attorney and client scale, the costs of the action as well as the costs of a number of interlocutory applications and the qualifying costs and attendance fees of the respondent’s experts. In respect of the action and the interlocutory applications, the costs of two counsel were allowed. It was contended on behalf of the appellant that these costs awards were not justified, particularly the punitive scale. Costs are in a trial court’s discretion. Absent any improper exercise of that judicial discretion, namely a fatal irregularity or misdirection or a startlingly inappropriate decision, a court of appeal will not interfere, even though it may take a different view on costs.[35] There are no grounds to interfere on appeal with the costs orders of the high court, which were motivated primarily by the appellant’s conduct, discussed extensively in this judgment.
[167] The costs in this court stand on a different footing. The appellant has obtained a substantial amelioration of the high court’s order. On the other hand, he has fallen well short of his objective of establishing that the waiver of maintenance was valid and that his estate showed no accrual. We thus think that in this court the parties should bear their own costs.
[168] Since the appellant will need time to realise assets in order to satisfy the accrual claim, we consider it appropriate, in terms of s 10 of the MPA, to defer payment of the accrual. The most likely source of capital will be from the sale of the Rondebosch property. A deferral until 1 December 2018 for this purpose should suffice. In the meanwhile the accrual award will attract interest at the prescribed rate. As indicated earlier, the respondent will be entitled to remain in occupation of the Rondebosch property rent-free until the accrual amount is paid to her.
[169] The following order is made:
[a] The appeal is upheld in part.
[b] Paragraphs 3, 4, 5, 6 and 7 of the order of the high court are set aside and substituted with the following (to avoid confusion, the paragraph numbering in the court a quo’s order is retained) :
‘3. Maintenance:
3.1. The defendant is directed to pay to the plaintiff maintenance as follows:
(a) R18 500 per month for one year as from 1 September 2016;
(b) R13 500 per month for one year as from 1 September 2017;
(c) R8500 per month as from 1 September 2018.
3.2. The obligation to pay maintenance as aforesaid shall endure until the plaintiff’s death or remarriage, whichever occurs first. The maintenance must be paid by way of debit order into such bank account as the plaintiff nominates from time to time and by not later than the first day of each month. The defendant shall be entitled to deduct from the amounts specified in 3.1(a) and 3.1(b) the amounts of maintenance already paid pending the appeal.
3.3. The amounts of maintenance specified in 3.1 above are expressed in nominal terms as at 1 September 2016. The amounts payable as from 1 September 2017 and 1 September 2018 respectively, and as from 1 September of each succeeding year, must be adjusted by the percentage change in the headline inflation rate (also known as the Headline Consumer Price Index) as notified by Statistics SA (or its equivalent) (‘the index’). Such percentage change shall for purposes of convenience be deemed to be equal to the latest index available from Statistics SA on the anniversary date.
4. The accrual in the defendant’s estate is held to be R 8 892 482.
5. The defendant shall pay to the plaintiff half of this amount, minus R70 000 in respect of the plaintiff’s net restitutionary obligation, ie a net amount of R4 376 241, by not later than 1 December 2018. Pending such payment, and as from 5 August 2016, interest shall run on the said net amount at the prescribed rate.
6. The Rondebosch property:
6.1 The plaintiff is ordered to transfer to the defendant her undivided half share in the property situated at 5 Woodlands Road, Rondebosch, Cape Town (the Rondebosch property) free of consideration.
6.2 Such transfer shall be effected as soon as is reasonably practicable after the date of the appeal judgment.
6.3 The reasonable costs of transfer shall be borne in equal shares by the parties.
6.4 Transfer shall be effected by attorneys appointed jointly by the parties, such appointment to be made within one month of the appeal judgment. If the parties cannot agree on the identity of such attorneys within the said one-month period, attorneys appointed by the President of the Law Society of the Cape of Good Hope shall be mandated by the parties to effect transfer.
6.5 The plaintiff shall be entitled to remain in occupation of the Rondebosch property rent-free until one month after the date on which payment of the amount in 5 above is effected. Any agreement for the sale of the property must be subject to this right of occupation.
7. The defendant may collect the movables specified in Exhibit 41 of the record from the Rondebosch property during the one-month period contemplated in 6.5 above.’
[c] The parties shall bear their own costs of appeal.
___________________
S A Majiedt
Judge of Appeal
___________________
O L Rogers
Acting Judge of Appeal
Majiedt JA (Saldulker and Dambuza JJA and Plasket AJA concurring):
[170] As foreshadowed in the joint judgment, this is my separate judgment on the unenforceability of the waiver of maintenance. The respondent advanced five grounds in support of her contention that the clause is per se against public policy. They are briefly as follows:
(a) First, that the clause seeks to exclude a court’s statutory power to award maintenance in terms of s 7 of the Divorce Act in the future.
(b) Second, that the clause seeks the future exclusion of the statutory right conferred by s 2 of the Maintenance of Surviving Spouses Act 27 of 1990, to claim maintenance from the appellant’s estate on the dissolution of the marriage by his death.
(c) Third, the clause seeks to exempt the appellant from the consequences of all and any misconduct by him, even that cognisable by the court in terms of s 7(2) of the Divorce Act.
(d
) Fourth, that it was a unilateral waiver applying to the respondent only, with no reciprocal waiver by the appellant.(e) Fifth, that while the donations in clauses 6 and 8 were ostensibly given as quid pro quo for the waiver, clause 3 of the contract provided, contrary to s 5(2) of the MPA[36], that such donations are to be taken into account as part of the accrual in the respondent’s estate.
The high court upheld all five grounds. In view of the conclusion I have reached, I do not intend dealing with all these grounds.
[171] For the reasons that follow, I am of the view that the waiver clause per se offends public policy, more particularly legal policy in the form of s 7 of the Divorce Act. At common law, a person had no right to maintenance after the dissolution of a marriage. A spouse’s right to maintenance upon divorce was introduced by s 10(1) of the Matrimonial Affairs Act 37 of 1953.[37] Section 10(1) of that Act empowered a court to make a maintenance award against the guilty spouse in favour of the innocent spouse. In that regard, the courts have generally held that a court had a general discretion to award maintenance to an innocent spouse.[38] Section 10(1) was replaced by s 7 of the Divorce Act which currently regulates maintenance for spouses after dissolution of a marriage. Sections 7(1) and (2) read as follows:
‘7 Division of assets and maintenance of parties
(1) A court granting a decree of divorce may in accordance with a written agreement between the parties make an order with regard to the division of the assets of the parties or the payment of maintenance by the one party to the other.
(2) In the absence of an order made in terms of subsection (1) with regard to the payment of maintenance by the one party to the other, the court may, having regard to the existing or prospective means of each of the parties, their respective earning capacities, financial needs and obligations, the age of each of the parties, the duration of the marriage, the standard of living of the parties prior to the divorce, their conduct in so far as it may be relevant to the break-down of the marriage, an order in terms of subsection (3) and any other factor which in the opinion of the court should be taken into account, make an order which the court finds just in respect of the payment of maintenance by the one party to the other for any period until the death or remarriage of the party in whose favour the order is given, whichever event may first occur.’
[172] There appears to be no decided cases on whether the prenuptial waiver of the right to maintenance upon dissolution of a marriage offends public policy. Professor Hahlo expressed the view that at common law such a waiver is contrary to public policy.[39] The learned author does not, however, cite any authority for the proposition that among the provisions that ‘would clearly be ineffective as being against public policy’ is a clause ‘depriving the courts of their statutory powers on divorce to award post-divorce maintenance to one of the spouses (s 7 of the Divorce Act)’. More recent works on matrimonial law are largely silent on this topic. Heaton et al,[40] in a discussion about the legality of terms in an antenuptial contract, list a number of terms which would be contrary to a rule of statutory law or contrary to good morals, public policy or the nature of marriage. The learned authors include as one of these a term that there will be no forfeiture of benefits after a divorce and they cite as authority the above passage from Hahlo. But they say nothing about a waiver such as the present one.[41] Waiver of the right to maintenance is discussed only in the context of waiver at the time of divorce. The learned authors also discuss waiver (at the time of divorce) of the right to apply for the rescission, suspension or variation of maintenance orders. As they correctly point out, the controversy and conflicting decisions on this last mentioned subject were finally settled by Schutte v Schutte[42] in this court. It is helpful to consider this aspect and to contrast it with the prenuptial waiver of maintenance.
[173] In Schutte this court held that the waiver by a spouse of the right to seek variation of a maintenance order in terms of s 8(1) of the Divorce Act is not against public policy. In a comprehensive analysis of the meaning and purview of the provisions of the old s 10(1) of the Matrimonial Affairs Act [43] and s 7(2) of the Divorce Act, Van Heerden JA emphasized that these provisions found application only upon divorce. The learned judge contrasted the legal position here with that in England where s 34(1) of the Matrimonial Causes Act of 1973 rendered waiver of the right to maintenance void as being ‘a provision purporting to restrict any right to apply to a court for an order containing financial arrangements’. In addition, such a waiver was also regarded in England as being contrary to public policy.[44]
[174] In my view there is a stark difference between waiver upon divorce of the right of a spouse to seek variation of a maintenance order, as envisaged in s 8(1), and a prenuptial waiver of maintenance. The main, compelling, difference is that at the time of divorce both spouses have full knowledge of their respective financial means and needs. That is not the case before the parties have married. It was pointed out in Schutte that, unlike in England, here a divorced spouse has no statutory remedy if no order for maintenance is granted upon divorce.[45] Section 7(2) was enacted (and before it, s 10 of the Matrimonial Affairs Act) to provide a statutory right to a spouse to obtain a maintenance order upon divorce. Public (legal) policy therefore establishes a statutory right to maintenance upon divorce. Such a right cannot be waived prenuptially – it would offend legal policy and hence public policy.
[175] The argument advanced on behalf of the appellant, that the waiver clause is valid, impermissibly adopted an overly narrow approach. It was contended that the clause did not breach any fundamental right in the Bill of Rights. That argument is fallacious. First, the respondent’s attack in her pleadings was not restricted to a constitutional challenge to the validity of the clause. Her contention that the impugned clause was inimical to constitutional values and infringed her rights to dignity and equality was but one of the strings to her bow. Hers was a wider challenge, namely that the clause was against public policy. And, second, while public policy is now rooted in the Constitution and its underlying values, it may sometimes extend beyond it.
[176] In the leading authority on the interrelationship between public policy, constitutional values and contractual autonomy, the Constitutional Court held that public policy is ‘to be determined by reference to the values that underlie our constitutional democracy as given expression by the provisions of the Bill of Rights’.[46] But, as Harms DP, correctly with respect, observed in Bredenkamp ‘[p]ublic policy and the boni mores are now deeply rooted in the Constitution and its underlying values. This does not mean that public policy values cannot be found elsewhere’[47] (emphasis added). Secondly, Barkhuizen concerned a direct constitutional challenge, namely the constitutionality of a time limitation clause in a short-term insurance policy. As stated, the respondent’s attack on the impugned clause went far wider, namely that it is against public policy. There is a significant difference between these two challenges. I endorse the observations made by Alkema J in this regard in Nyandeni Local Municipality:
‘In my respectful view, there is a difference in approach to an attack on the constitutionality of a term of contract on the ground of it being inconsistent with the Constitution, on the one hand; and on the other hand, an attack on the validity or enforceability of a contract or a term thereof on the ground of it being in conflict with public policy. In the latter case the concept of public policy is informed by the underlying values and principles of the Constitution, and it is in this sense only that the constitutional order is relevant. In a direct constitutional attack, the constitutional right must first be identified and secondly such right must be then found to be limited by “a law of general application”.’ [48]
[177] The appellant’s counsel understandably laid much emphasis on the trite principle that contracts must be honoured. But, as Harms DP stated in Bredenkamp, pactum sunt servanda is no holy cow. The determination whether an agreement offends public policy entails the weighing up of competing values and pactum sunt servanda is but one such value.[49] Agreements that are contrary to public policy were not recognized at common law. As this court stated in Sasfin (Pty) Ltd v Beukes:[50]
‘Agreements which are clearly inimical to the interests of the community, whether they are contrary to law or morality or run counter to social or economic experience, will accordingly, on the grounds of public policy, not be enforced’.[51]
As Cameron JA said in Brisley v Drotsky, public policy is now infused with the normative values in the Bill of Rights. Moreover, contractual autonomy is not the antithesis of fundamental rights:
‘[C]ontractual autonomy is part of freedom. . . [C]ontractual autonomy informs also the constitutional value of dignity.’[52]
The traditional view of the sanctity of contract has over the last few years undergone a profound realignment in view of the Bill of Rights.[53]
[178] Related to this is the well-established common law principle that an aggrieved person has the right to seek the assistance of a court of law and that a term in a contract which deprives a party of the right to approach a court for redress is contrary to public policy.[54] Commonwealth jurisdictions appear to favour the approach that prenuptial agreements as well as agreements concluded stante matrimonio and post-nuptially cannot oust a court’s statutory power through a waiver of maintenance.[55] For present purposes, however, it is in my view sufficient to find that the impugned clause offends public policy as it is inimical to the legal policy regarding maintenance, encapsulated in s 7 of the Divorce Act. Such a finding accords with well-established sound legal precedent developed over decades in this country. No reasons are suggested for departing from this approach in favour of the novel ‘nuanced and enlightened approach’ of some Commonwealth jurisdictions, as espoused by Rogers AJA in his separate concurrence. The latter approach may commend itself in respect of postnuptial agreements, but I express no firm view on it.
[179] In Claassens v Claassens[56], Didcott J was confronted with the question whether a waiver of the right to apply for an increase in maintenance, contained in a divorce settlement, offends public policy. The learned Judge held that it did not. With reference to Schierhout, Didcott J pointed out that ‘public policy frowns on the transaction only when the particular remedy that is waived is one it wants retained. What offends public policy outside the Schierhout rule, in other words, is not the exclusion of the court’s jurisdiction per se, but its exclusion from matters which public policy insists on keeping justiciable’[57] (emphasis added).This is the approach which I think should be followed in this case.
[180] The importance of marriage as a social institution with profound significance, not only to the spouses concerned, but also to society at large, was recognized by the Constitutional Court in Dawood.[58] The Court observed that ‘[t]he celebration of a marriage gives rise to moral and legal obligations, particularly the reciprocal duty of support placed upon spouses . . . These legal obligations perform an important social function . . . Importantly, the community of life establishes a reciprocal and enforceable duty of financial support between the spouses . . .’[59]
[181] In EH v SH, this court stated that s 7 of the Divorce Act was enacted to, amongst others, alleviate the potential iniquitous situation where a wife had spent her active economic years caring for the children and running the joint household.[60] There can be little doubt that the institution of marriage and the traditional role of the ‘stay at home’ wife have changed significantly in recent times, but the impact of prenuptially waiving one’s right to maintenance upon dissolution of the marriage must not be underestimated. During argument much was made of the fact that the respondent had waived her right to maintenance on the basis that she would receive the donations outlined in the contract. In my view that consideration does not save the impugned clause from invalidity. A spouse would have no idea prior to marriage how long the marriage would last, what her or his needs and means may be at the time of dissolution of the marriage and, generally, what the future holds. In that regard, she or he is in a markedly different position than a spouse who enters into such an agreement stante matrimonio. Even a newlywed spouse would have a much better understanding of what the future holds and of the parties’ respective means and needs, both present and in future. Before the marriage, a prospective spouse would have no idea whether the donations would, absent any maintenance, be adequate to meet her or his future needs upon divorce.
[182] In conclusion: the waiver clause is contrary to legal policy (s 7 of the Divorce Act) and therefore offends public policy.
____________________
S A Majiedt
Judge of Appeal
Rogers AJA:
[183] I agree with Majiedt JA that the maintenance waiver is not enforceable in the present case but my reasons differ from his. He finds that a prenuptial waiver of maintenance is by its nature contrary to public policy. This entails that the waiver is void ab initio (Sasfin (Pty) Ltd v Beukes 1989 (1) SA 1 (A) at 18G-I). The public policy on which he bases his conclusion is the policy reflected in ss 7(1) and (2) of the Divorce Act. I agree with his assessment of the policy. However, before one concludes that a prenuptial waiver is contrary to public policy and void, it is necessary to interpret the legislation to determine whether it does not already accommodate the policy concerns. If it does, there is no need to brand a prenuptial waiver as per se contrary to public policy. To do so would be contrary to the legislative scheme. My conclusion is that the legislation does indeed provide a mechanism to address my colleague’s policy concerns, namely by giving the divorce court an overriding discretion to disregard the waiver.
[184] The provisions of ss 7(1) and (2) have been quoted in my colleague’s separate judgment. The appellant’s counsel argued that the ‘written agreement’ contemplated in s 7(1) could include a prenuptial agreement and that, whenever there is a written agreement regulating maintenance, the court’s power to make an order in terms of s 7(2) is excluded. The respondent’s counsel submitted, by contrast, that the ‘written agreement’ in question is confined to an agreement concluded at the time of divorce.
[185] The expression ‘written agreement’ in s 7(1) is not limited in point of time, so the respondent’s argument would require one to read words into the subsection. Furthermore, if one regards only prenuptial contracts as falling outside the ambit of s 7(1), a maintenance waiver in an agreement concluded after the commencement of the marriage would be valid even though it was executed before the breakdown of the marriage or a long time before the divorce. If further limitations must be read into the section in respect of postnuptial contracts, it is not clear to me where the dividing line would be.
[186] I thus do not think that the limitation advanced by the respondent is justified. This, however, does not lead to the result for which the appellant’s counsel contended. In terms of s 7(2), the court’s power to make an order under that subsection exists in any case where an order has not been made in terms of s 7(1). Section 7(1) provides that a court ‘may’, not ‘must’, make an order in accordance with a written agreement of the kind contemplated. If a court considers that there is good reason not to give effect to the written agreement regarding maintenance, it may refrain from doing so and can then proceed to make an order in terms of s 7(2).
[187] Read together, ss 7(1) and (2) do not prohibit an agreement by which a spouse waives her right to maintenance in return for gifts but they do explicitly accord to the court a discretion either to give effect to the agreement in terms of s 7(1) or to award maintenance in terms of s 7(2). The very circumstance that the court has a statutory power to override the agreement shows that an agreement cannot override the statutory power. This flows inevitably from a proper interpretation of the statutory provisions though it is supported by considerations of policy. As was said in Ritch and Bhyat v Union Government (Minister of Justice) 1912 AD 719, to allow such a waiver would be ‘to defeat the provisions of an enactment intended on general and public grounds to be peremptory and binding on all concerned’ (at 735). However, the fact that the overriding statutory power cannot be ousted by contract does not lead to the conclusion that the parties’ endeavour at the contractual ordering of maintenance is contrary to public policy.
[188] In the modern era there is much to be said for the view that our law should encourage the private ordering of the financial consequences of divorce and follow the approach which currently prevails in England, Canada and elsewhere, with such adaptations as may be appropriate to our country’s circumstances. The leading decision in England is Radmacher v Granatino [2010] UKSC 42, in which, unusually, the panel comprised nine justices. By its decision, the Supreme Court swept away the distinctions previously drawn in England between prenuptial and postnuptial agreements, holding that an agreement of either kind is a factor to be taken into account in determining whether the court should exercise its overriding power to award maintenance. In the leading judgment (carrying the support of eight justices) the court said this:
‘The court should accord respect to the decision of a married couple as to the manner in which their financial affairs should be regulated. It would be paternalistic and patronising to override their agreement simply on the basis that the court knows best.’ (Para 78)
The court recognised, however, that the scope for unfairness in the enforcement of a nuptial agreement was greater where the agreement was concluded prenuptially than where it was concluded after the marriage had broken down:
‘Where the ante-nuptial agreement attempts to address the contingencies, unknown and often unforeseen, of the couple's future relationship there is more scope for what happens to them over the years to make it unfair to hold them to their agreement. The circumstances of the parties often change over time in ways or to an extent which either cannot be or simply was not envisaged. The longer the marriage has lasted, the more likely it is that this will be the case’. (Para 80)
But, as the court had earlier remarked, there is no fundamental distinction between the case of a prenuptial and a postnuptial contracts:
‘Nuptial agreements made just after the wedding are not unknown and likely to become more common if the law distinguishes them from ante-nuptial agreements.’ (Para 57)
[189] Even in the earlier judgment of the House of Lords in Hyman v Hyman [1929] AC 601, based to some extent on outmoded views of marriage and divorce, the majority did not hold that a marriage agreement involving a waiver of maintenance was contrary to public policy, only that the statutory power to order maintenance could not be overridden. The majority acknowledged that the agreement between the parties might be relevant to the exercise of the statutory power.
[190] In Versteegh v Versteegh [2018] EWCA Civ 1050 King LJ said that Radmacher represented a ‘sea change’ in the English law’s approach to pre-marital agreements. Not only are such contracts no longer contrary to public policy: ‘Where a party has a full appreciation of its implications, the court should now give effect to such an agreement, unless it would be unfair to do so’ (para 44). In BN v MA [2013] EWHC 4250 (Fam), Mostyn J stressed the importance of the principle of personal autonomy, a consideration he thought of particular importance where the parties are ‘sophisticated, highly intelligent and have the benefit of the best legal advice that money can buy’. Where, in those circumstances, they have ‘thrashed out’ an agreement, ‘heavy respect’ should be accorded to the agreement, particularly where it seeks to protect premarital property (para 28). Few could quibble, I would have thought, with the enforcement of the antenuptial contract in H v H [2016] EWFC B81, where the marriage lasted only twelve weeks.
[191] In Canada the approach was initially laid down in Pelech v Pelech [1987] 1 SCR 801 and is now to be found, on account of intervening legislative changes, in Miglin v Miglin 2003 SCC 24; [2003] SCR 303. The essence of the Pelech approach was described in Miglin as being
‘that a court will not interfere with a pre-existing agreement that attempts fully and finally to settle the matter of spousal support as between the parties unless the applicant can establish that there has been a radical and unforeseen change in circumstances that is causally connected to the marriage.’
[192] In Miglin the statutory provision under consideration was s 15.2(4)(c) of the Divorce Act 1985, which provided that, in the making of an award of spousal maintenance, the court should take into account, among other matters, ‘any agreement or arrangement relating to support of either spouse’. The majority’s approach is reflected in the following passage at 333-334:
‘[T]he language and purpose of the 1985 Act militate in favour of a contextual assessment of all the circumstances. This includes the content of the agreement, in order to determine the proper weight it should be accorded in a s.15.2 application. In exercising their discretion, trial judges must balance Parliament’s objective of equitable sharing of the consequences of marriage and its breakdown with the parties’ freedom to arrange their affairs as they see fit. Accordingly, a court should be loathe to interfere with a pre-existing agreement unless it is convinced that the agreement does not comply substantially with the overall objectives of the Divorce Act. This is particularly so when the pre-existing spousal support agreement is part of a comprehensive settlement of all issues related to the termination of the marriage.’
[193] The court proceeded to formulate a two-stage process in assessing whether to give effect to the pre-existing spousal support agreement. At stage 1 the court considers the circumstances prevailing when the agreement was negotiated to determine whether there is any reason to discount it (a power imbalance, oppression, other conduct falling short of unconscionability, the duration of negotiations, the presence or absence of professional advice, the extent to which the agreement at the time of its conclusion was in substantial compliance with the objectives of the Divorce Act). At stage 2 the court assesses the extent to which the agreement still reflects the original intention of the parties and the extent to which it is still in compliance with the objectives of the Divorce Act. A certain degree of change is always foreseeable by spouses when they conclude an agreement, leading the majority to say the following (at 353):
‘Although we recognize the unique nature of separation agreements and their differences from commercial contracts, they are contracts nonetheless. Parties must take responsibility for the contract they execute as well as for their own lives. It is only where the current circumstances represent a significant departure from the range of reasonable outcomes anticipated by the parties, in a manner that puts them at odds with the objectives of the Act, that the court may be persuaded to give the agreement little weight.’
[194] Although Pelech and Miglin concerned postnuptial agreements, the same two-stage approach is applied by Canadian courts to all pre-existing agreement, whether prenuptial or postnuptial. Prenuptial waivers of spousal maintenance appear to be quite common in Canada. Some of the cases where Miglin was applied to prenuptial waivers are discussed by Prof Carol Rogerson in her article ‘Spousal Support Agreements and the Legacy of Miglin’ 31 Canadian Family Law Quarterly 13-70.[61]
[195] Since section 7(1) of our Divorce Act refers not only to agreements regarding maintenance but also agreements regarding division of property, it is not out of place to referred to a more recent judgment of the Supreme Court of Canada in Hartshorne v Hartshorne [2004] 1 SCR 550, 2004 SCC 22 (CanLII) where the two-stage approach in Miglin approach was applied to a prenuptial property division agreement governed by the Family Relations Agreement of British Columbia. In para 39 Bastarache J, who delivered the majority judgment, said the following:
‘This Court has not established, and in my opinion should not establish, a “hard and fast” rule regarding the deference to be afforded to marriage agreements as compared to separation agreements. In some cases, marriage agreements ought to be accorded a greater degree of deference than separation agreements. Marriage agreements define the parties’ expectations from the outset, usually before any rights are vested and before any entitlement arises. Often, perhaps most often, a desire to protect pre acquired assets or an anticipated inheritance for children of a previous marriage will be the impetus for such an agreement. Separation agreements, by contrast, purport to deal with existing or vested rights and obligations, with the aggrieved party claiming he or she had given up something to which he or she was already entitled with an unfair result. In other cases, however, marriage agreements may be accorded less deference than separation agreements. The reason for this is that marriage agreements are anticipatory and may not fairly take into account the financial means, needs or other circumstances of the parties at the time of marriage breakdown.’
[196] In Australia the enforceability of prenuptial agreements regulating, inter alia, post-divorce spousal support is now regulated by Part VIIIA of the Family Law Act 1975, inserted into the principal Act in 2000. (For a discussion of these provisions, see inter alia Hoult v Hoult [2013] Fam CAFC. 109.) Prior to such statutory regulation, the Australian courts had arrived at a position not unlike that espoused in Radmacher: the court would have regard to, but was not bound to implement, a prenuptial contract dealing with the division of property or the payment of post-divorce spousal support.[62]) In New Zealand, s 128 of the Family Proceedings Act 1980 appears to perform a similar function to the Australian legislation (cf Ward v Ward [2009] NZSC 125; [2010] 2 NZLR 31). In these two countries, as far as I can ascertain, prenuptial agreements regarding maintenance are not invalid but there are safeguards against unfairness when it comes to enforcement at the time of divorce.
[197] Sections 7(1) and (2) of our Act lend themselves admirably to an interpretation allowing us to follow the nuanced and enlightened approach prevailing in England, Canada and elsewhere– to operate with the statutory scalpel rather than the common law cutlass. A South African court, considering a claim for maintenance in the face of a prenuptial or postnuptial agreement containing a maintenance waiver (or other maintenance provisions inconsistent with a claim advanced by a spouse at the divorce hearing), should consider a range of factors in deciding whether to award maintenance or (as was done in Radmacher and in several Canadian cases) to hold the parties to the contract. The sorts of factors to be taken into account are likely to include most of those mentioned in the leading English and Canadian decisions. This interpretation not only accords with the plain language of the sections but seems to me to give better effect to constitutional values – it eschews paternalistic thinking and promotes party autonomy while at the same time giving the court a generous jurisdiction to prevent unfair outcomes.
[198] In essence, the competing considerations which are engaged in assessing prenuptial contracts relating to post-divorce division of property and spousal support are autonomy and protection. Both are relevant considerations. My colleague’s approach promotes protection to the complete exclusion of autonomy. The appellant’s fallacious argument promotes autonomy to the complete exclusion of protection. The approach I advocate allows both considerations to play a role in a careful, fact-specific enquiry. My approach does not compromise any of the policy considerations which concern my colleague. And it keeps us in step with leading Commonwealth jurisdictions.
[199] In the present case it is unnecessary to delineate with greater precision the test to be applied by a South African court when deciding whether to give effect to a prenuptial contract of the kind contemplated in s 7(1). Whatever approach were adopted, the result in the present case would be that the provisions of the antenuptial contract relating to maintenance should be disregarded and that the matter should be assessed afresh in the light of the considerations listed in s 7(2). The main factor which impels me to this conclusion is the lengthy period that has elapsed between the conclusion of the antenuptial contract and the divorce (more than 25 years), with all the unforeseen changes in circumstances which have occurred in that time. Other relevant factors would be that the respondent was placed under some pressure to accept the terms of the antenuptial contract, that she did not have independent legal advice from a South African lawyer and that her emotional condition was compromised by the stress of her recent law examinations and the impending birth of her son.
[200] This conclusion does not imply criticism of the appellant. He was under pressures of his own, in particular his desire to be married to the mother of his child, his concern that she might marry another man if things were not promptly arranged and his concern to avoid a repeat of the costly divorce which had terminated his first marriage. If, as both prospective spouses foresaw as a possibility, their marriage was short-lived, the antenuptial contract might – from the perspective of maintenance – have been quite generous to the respondent. She was young, well-qualified and would probably not have succeeded in a claim for any or substantial maintenance whereas she would have got to keep the donations given as quid pro quo for the waiver.
[201] Certain of the high court’s findings in connection with the conclusion of the antenuptial contract were not warranted but I find it unnecessary to deal with them at length. I simply record that the high court’s description of the respondent as ‘gullible and naive’ was a demeaning description of an intelligent woman in her mid-twenties. Also unjustified was the high court’s statement, more than once in its judgment, that the appellant ‘planned this divorce even before he concluded this marriage’. The appellant did not marry the respondent with the intention of divorcing her. At no stage during the parties’ sometimes troubled relationship was it the respondent who wanted a divorce. The antenuptial contract contained provisions catering for the possibility of a dissolution of the marriage. That is not unusual or wrong.
___________________
O L Rogers
Acting Judge of Appeal
APPEARANCES
For Appellant |
M Konstantinides SC (with him F Gordon-Turner) |
|
Instructed by: |
|
Petra Visser Attorneys, Johannesburg McIntyre van der Post Inc, Bloemfontein |
|
|
For Respondent |
J G Dickerson SC (with him J S Anderson) |
|
Instructed by: |
|
Miller du Toit Cloete Inc, Cape Town Phatshoane Henney Attorneys, Bloemfontein |
[1] R v Dhlumayo & another 1948 (2) SA 677 (A) at 705 –706; Protea Assurance Co Ltd v Casey 1970 (2) SA 643 (A) at 648E; Matlou v Makhubedu 1978 ( 1) SA 946 (A) at 950 A-E.
[2] Brookstein v Brookstein [2016] ZASCA 40; 2016 (5) SA 210 (SCA) para 19.
[3] Harms Civil Procedure in the Superior Courts, B – 246 (Issue 53).
[4] Makate v Vodacom (Pty) Ltd [2016] ZACC 13; 2016 (4) SA 121 (CC) para 38
[5] Powell & Wife v Streatham Nursing Home [1935] 243 AC 265.
[6] Section 4 of the MPA; Brookstein v Brookstein, fn 2, paras 16 and 19.
[7] For present purposes the exclusions outlined in the proviso bear no relevance. The provisions of s 5(2) of the MPA have been cited in footnote 1 above.
[8] This stipulation accords with s 5(1) of the MPA.
[9] These exclusions originate from s 4(1)(b)(ii) of the MPA.
[10] MB v DB [2013] ZAKZDHC 33; 2013 (6) SA 86 (KZD) para 39.
[11] Sections 23 and 24 of the Matrimonial Causes Act of 1973, which have not been affected by the subsequent enactment of the Matrimonial and Family Proceedings Act of 1984 and the Family Law Act of 1996. The applicable rules are rules 73 – 76 of the Matrimonial Causes Rules of 1977.
[12] Livesey (formerly Jenkins) v Jenkins [1984] UKHL 3; [1985] 1 All ER 106 (HL).
[13] Livesey v Jenkins above, at 115.
[14] Prest v Petrodel Resources Limited & others [2013] UKSC 34 para 4.
[15] NG v SG [2011] EWHC 3270 (Fam).
[16] B v B [2014] ZASCA 137; 25 September 2014 paras 39 and 40.
[17] NG v SG, fn 15.
[18] AM v JM (3954/10); 2011 JDR 0091 (WCC) para 43.
[19] MB v DB, fn 10.
[20] B v B, fn 16, para 33.
[21] Eagle Star Insurance Co Ltd v Willey 1956 (1) SA 330 (A) at 334A – 335F; Van Zyl NO v Kiln Non-Marine Syndicate No 510 of Lloyds of London [2002] ZASCA 120; [2002] 4 All SA 355 (SCA) para 7.
[22] Minister of Law and Order v Monti [1994] ZASCA 139; 1995(1) SA 35 (A) at 40C – D.
[23] AA Onderlinge Assuransie-Assosiasie Bpk v De Beer 1982 (2) SA 603 (A) at 614G.
[24] Cooper & another NNO v Merchant Trade Finance Ltd 2000 (3) SA 1009 (SCA) at 1027-8.
[25] NG v SG, fn 15, para 16.
[26] M v M (26868/14) [2016] ZAGPJHC 387 (10 August 2016).
[27] Ibid, para 3.
[28] De Kock v Jacobson and another 1999 (4) SA 346 (W) at 350G; Mcintosh v Mcintosh [2011] ZAFSHC 116 paras 19-22; Elesang v PPC Lime Limited and others 2007 (6) SA 328 (NC) para 20; Kirkland v Kirkland 2006 (6) SA 144 (C) paras 82 and 83; Government Employees Pension Fund v Naidoo 2006 (6) SA 304 (SCA).
[29] Published in Government Gazette No 32005.
[30] Yarona Healthcare Network (Pty) Ltd v Medshield Medical Scheme [2017] ZASCA 116; 2018 (1) SA 513 (SCA) para 48.
[31] African Diamond Exporters (Pty) Ltd v Barclays Bank International Ltd 1978 (3) SA 699 (A) at 713.
[32] Kudu Granite Operations (Pty) Ltd v Caterna Ltd 2003 (5) SA 193 (SCA) para 28.
[33] R466 806 (Caledon Street Guest House loan account) + R858 952 (net recoverable component of Mont du Toit loan account). The supposed loan account of R1.55 million in Blouvlei must be left out of account for reasons previously explained.
[34] R1 245 028 (loan claim against Gamsberg) – R269 004 (loan owed to Schoongezicht.
[35] Ward v Sulzer 1973 (3) SA 701 (A) at 707A.
[36] Section 5(2) of the MPA reads: ‘In the determination of the accrual of the estate of a spouse a donation between spouses, other than a donation mortis causa, is not taken into account either as part of the estate of the donor or as part of the estate of the donee’.
[37] See: Strauss v Strauss 1974 (3) SA 79 (A) at 93 H.
[39] H R Hahlo, The South African Law of Husband and Wife, 5 ed (1985) at 259.
[40] Heaton et al The Law of South Africa (2 ed) vol 16.
[41] Ibid para 116.
[42] Schutte v Schutte 1986 (1) SA 872 (A) at 883-884.
[43] There had been conflicting decisions on whether under s 10(1) a waiver of the right to seek an increase of an amount agreed for maintenance in a consent paper was against public policy.
[44] Hyman v Hyman [1929] AC 601; All ER 245; Jessel v Jessel [1979] 3 All ER 645 at 649.
[45] Schutte v Schutte, fn 48, at 884A.
[46] Barkhuizen v Napier [2007] ZACC 5; 2007 (5) SA 323 (CC); 2007 (7) BCLR 691 (CC) para 29.
[47] Bredenkamp v Standard Bank of SA Ltd [2010] ZASCA 75; 2010 (4) SA 468 (SCA); 2010 (9) BCLR 892 (SCA); [2010] 4 All SA 113 (SCA) para 39; see also Nyandeni Local Municipality v MEC for Local Government and Traditional Affairs 2010 (4) SA 261 (ECM) para 79.
[48] Ibid para 72.
[49] Bredenkamp v Standard Bank paras 37-38.
[50] Sasfin (Pty) Ltd v Beukes 1989 (1) SA 1 (A).
[51] Sasfin (Pty) Ltd v Beukes at 8C – D.
[52] Brisley v Drotsky 2002 (4) SA 1 (SCA) para 94; see also: Barkhuizen v Napier para 30.
[53] See, for example, the discussion in Woolman et al Constitutional Law of South Africa 2 ed, at 31 – 37 and 31 – 130.
[54] Schierhout v Minister of Justice 1925 AD 417 at 424.
[55] See, inter alia, in the United Kingdom: Hyman v Hyman [1929] AC 601 but compare recent developments in Radmacher v Granatino [2010] UKSCC 42 at para 52; Canada: Miglin v Miglin 2003 SCC 24; [2003] 1 SCR 303; Australia: Brooks v Burns Philp Trustee Co Ltd [1969] HCA 4.
[56] Claassens v Claassens 1981 (1) SA 360 (N).
[57] Claassens v Claassens at 366G – H.
[58] Dawood & another v Minister of Home Affairs & others; Shalabi & another v Minister of Home Affairs & others; Thomas & another v Minister of Home Affairs & others [2000] ZACC 8; 2000 (3) SA 936 (CC).
[59] Ibid paras 31 and 33.
[60] EH v SH [2012] ZASCA 19; 2012 (4) SA 164 (SCA) para 12.
[61] See Loy v Loy [2007] OJ No 4274 and Frazer v van Rootselaar 2006 BCCA 198 (CanLII) (where Miglin challenges failed – discussed by Rogerson at p 52); and Varney v Varney 2008 NBQB 389 (CanLII), Jenkins v Jenkins 2008 MBQB 271 (CanLII) and M(L) v M(I) [2007] NJ No 379, 2007 NLUFC 29 (where Miglin challenges succeeded – discussed by Rogerson at p 60). See also Charles v Charles 1991 BCSC 551 (CanLII); Small v Small 1993 BCSC 1709 (CanLII); Segal v Qu 2001 ONSC 28201 (CanLII).
[62] Belinda Fehlberg and Bruce Smyth ‘Pre-Nuptial Agreements for Australia: Why Not? (2000) 14 Australian Journal of Family Law 80 at 81.