Trust assets and accrual claims at divorce: The SCA opens the door

August 1st, 2017

By Bradley Smith

In the article ‘Community of property and accrual sharing in terms of the Matrimonial Property Act, 1984 – Part 2’ (1985 (Feb) DR 59) Professor Andreas van Wyk cautioned that the Matrimonial Property Act 88 of 1984 (the MPA), which introduced the accrual system into South African law only three months earlier, seemed ‘deficient in protective measures against fraudulent alienations of assets aimed at defeating the other spouse’s equalization [accrual] claim’. In this regard, Prof van Wyk specifically mentioned alienations of property to a discretionary trust as a manner to engage in ‘preventative estate planning prior to divorce’ (at p 60). The spate of High Court judgments over the past decade that have dealt with the question as to whether the (value of) assets of a trust may be taken into account for the purposes of assessing the accrual of a spouse’s estate, testifies to the prescience of this statement. Finally, as of March of this year, we have a definitive answer to this question, by virtue of the Supreme Court of Appeal’s (SCA) judgment in REM v VM 2017 (3) SA 371 (SCA). The purpose of this article is to shed light on some aspects of this case and its vital importance for legal practitioners.

The debate over the past decade

The 2006 SCA judgment of Badenhorst v Badenhorst 2006 (2) SA 255 (SCA) provides a convenient point of departure. In this case, the SCA held that the value of the assets of a trust could be taken into account for the purposes of a redistribution of assets order (in terms of
s 7(3) to (6) of the Divorce Act 70 of 1979). This was because ‘the terms of the trust deed’ and ‘the evidence of how the affairs of the trust were conducted during the marriage’ proved that the respondent had de facto ‘controlled the trust and but for the trust would have acquired and owned the [trust’s] assets in his own name’ (at para 9, of the Badenhorst case) (the control test).

For present purposes two important principles may be distilled from the Badenhorst case:

  • First, the rationale for taking the trust’s asset values into account was that the respondent had had ‘full control of the trust’ and, in so doing, had ‘paid scant regard to the difference between trust assets and his own assets’ (at para 11). He had, therefore, violated the ‘core idea’ of the trust, which is to maintain a separation between the control (qua trustee) of trust property and the enjoyment of the benefits derived from such control (Land and Agricultural Bank of South Africa v Parker and Others 2005 (2) SA 77 (SCA) at paras 19 and 22). In short, the application of the ‘control test’ showed that the appellant had abused the trust by using it as his alter ego.
  • Secondly, and closely linked hereto, there was an important reason why the court was constrained to taking the value of the trust assets into account (as opposed to ordering the redistribution of the trust assets themselves): This was because, in contrast to the situation where a ‘trust’ is held to be a sham (due to non- [or simulated] compliance with the essential elements for the creation of a trust), the trust property had been validly transferred to the trustees of a legally created trust. The result was that they had obtained the ‘character’ of trust assets (see Van Zyl and Another NNO v Kaye NO and Others 2014 (4) SA 452 (WCC) at paras 16 – 23).

Bearing these principles in mind, the value of the trust assets could be taken into account to determine the extent of the redistribution order. In so doing, it is my view that the Badenhorst case entailed a piercing of the trust’s ‘veneer’ (see BS Smith ‘Statutory discretion or common law power? Some reflections on “veil piercing” and the consideration of (the value of) trust assets in dividing matrimonial property at divorce – Part One’ (2016) 2 Journal for Juridical Science 68 at 83 – 89 (Smith 2016), but compare to the  Kaye case at paras 23 and 24).

In the wake of the Badenhorst case, two schools of thought developed regarding the possibility of taking the value of trust assets into account in the context of accrual claims. The first, as exemplified by case law such as RP v DP and Others 2014 (6) SA 243 (ECP), was that the Badenhorst case involved the exercising of a common law power that had been transplanted from company law (ie, the principles pertaining to ‘piercing the corporate veil’) into trust law. As such, Alkema J opined that it was ‘fallacious’ to argue that the court’s discretion to take the value of the assets of an alter ego trust into account was derived from (and restricted to) marriages to which the discretionary powers conferred by s 7(3) to (6) of the Divorce Act applied (at para 56). In the result, the common law power could also be exercised in the accrual context.

The second school of thought is encapsulated in the approach adopted by Ploos van Amstel J in MM and Others v JM 2014 (4) SA 384 (KZP). In essence, the court held that there was ‘a fundamental difference’ between redistribution orders and accrual claims because in contrast to the former, ss 3, 4 and 5 of the MPA simply provided for a factual, mathematical calculation and did not permit a court ‘to make an assessment of what it deems to be “just”’ (at para 12 in the MM case). The MPA did, therefore, not permit any asset that did not form part of a spouse’s estate to be taken into account in calculating the accrual claim (at para 19 in the MM case).

The next major development was occasioned by the SCA’s judgment in WT and Others v KT 2015 (3) SA 574 (SCA). Although this matter dealt with a marriage in community of property, the court dealt two implicit, but decisive, blows to the possibility of trust assets being considered in the accrual context. The first, was the court’s finding that a crucial distinguishing feature between a marriage in community of property and the Badenhorst case was the absence of a ‘comparable discretion as envisaged in s 7(3) of the Divorce Act, to include the assets of a third party in the joint estate’ (at para 35 of the WT matter). Academic commentators at the time correctly concluded that this finding by implication extended to accrual claims and thus endorsed the approach in the MM case (see F du Toit ‘South Africa – Trusts and the patrimonial consequences of divorce: Recent developments in South Africa’ (2015) Journal of Civil Law Studies 654 at 696 and 697).

Secondly, the court held that an aggrieved spouse would only have ‘standing to challenge the management’ of a trust (in an attempt to have its veneer pierced) if a fiduciary responsibility was owed to that spouse because he or she was a trust beneficiary or a third party who had transacted with the trust. It goes without saying that this decision placed aggrieved divorcing spouses whose marriages did not fall within the ambit of s 7(3) of the Divorce Act and/or who were neither trust beneficiaries nor third parties who had contracted with the trust, in an invidious position. In effect, the message conveyed was that trustee-spouses whose marriages were subject to the accrual system were free to engage in unscrupulous ‘divorce planning’ by utilising an alter ego trust to minimise their accrual liability at divorce (see Smith 2016 at 85 and BS Smith ‘Statutory discretion or common law power? Some reflections on “veil piercing” and the consideration of (the value of) trust assets in dividing matrimonial property at divorce – Part Two’ (2017) 1 Journal for Juridical Science 1 at 7 – 14 (Smith 2017)).

The latest judgment of the SCA in REM v VM has, however, rectified this unsavoury state of affairs.

The REM case

In this case the parties had married and divorced one another on no less than three occasions. When the final divorce order was granted in 2011, the respondent’s patrimonial claim was, by agreement that was made an order of court, postponed sine die. It will suffice for the purposes of this article to accept the SCA’s finding that the exclusion clause contained in the parties’ antenuptial contract did not exclude the assets of two trusts (the S Trust and the CPB Trust) for accrual purposes. As such, the major question to be answered was ‘whether [these assets] legitimately form part of the assets of these trusts and do not form part of the appellant’s estate, for purposes of the accrual system’ (at para 10). This question, quite correctly, required the court to determine whether these trusts were merely the alter egos of the appellant. Although the SCA found that a conspectus of the evidence did not support the latter finding, two aspects of the case are of crucial importance for the purposes of this article.

First, the SCA confirmed the distinction between sham trusts and instances where a litigant sought to ‘pierce the veneer’ of a trust as the alter ego of a trust founder or trustee: While the former implied that a valid trust never existed, the latter entailed a remedy aimed at combating the abuse of an existing trust. This remedy could be styled as –

‘an equitable remedy in the ordinary, rather than technical, sense of the term; one that lends itself to a flexible approach to fairly and justly address the consequences of an unconscionable abuse of the trust form in given circumstances. It is a remedy that will generally be given when the trust form is used in a dishonest or unconscionable manner to evade a liability, or avoid an obligation’ (at para 17) (my italics).

Secondly, the SCA emphatically rejected the finding in the WT case that an aggrieved spouse, who was neither a beneficiary of the trust, nor a third party who had transacted with it, had no standing to impugn the management of a trust because no fiduciary duty was owed to such a spouse. In fact, as the SCA pointed out in the REM case, the fact that no fiduciary duty was owed to such a spouse was irrelevant. The critical issue instead was whether the spouse was seeking to advance a patrimonial claim that the other spouse had attempted to prejudice by administering a trust in such a manner as to amount to an unconscionable abuse of the trust form. Of crucial importance, the SCA correctly held that in marriages involving the accrual system, this would occur where a trustee-spouse transferred personal assets to a trust and dealt with them as if they were trust assets in a fraudulent or dishonest attempt to conceal them and thus prejudice the aggrieved spouse’s accrual claim. In these circumstances, trust assets could ‘be used to calculate the accrual’ of the errant trustee-spouse’s estate and to ‘satisfy any personal liability of [that spouse] to make payment to the [aggrieved spouse]’ (at paras 19 and 20).

Implications and practical relevance

The judgment in the REM case makes it clear that the asset values of an alter ego trust may now be taken into account in marriages that are subject to the accrual system. Although the court did not elaborate on the precise legal basis for doing so, it is my view that the judgment by implication endorses the sentiment expressed in the RP case to the effect that the power to pierce the veneer of an alter ego trust stems from the common law and exists independently of the Divorce Act or the MPA. As such, it may in principle be exercised in the context of all matrimonial property regimes. However, as I have pointed out in another publication (see Smith 2016 at 85 – 89) the fact that this power exists does not necessarily permit it to be exercised. Doing so depends on whether a nexus exists between the power and the applicable matrimonial property regime.

As far as future litigation is concerned, practitioners should note that two distinct processes are involved. The first step is to prove that one is dealing with a trust that is merely the alter ego of one of the spouses. This question may be answered with reference to the so-called ‘control test’ as enunciated in the Badenhorst case (op cit). Once this test has been complied with, the next step is to determine whether the applicable matrimonial property regime imposes an obligation on the divorcing spouses, which one of them has attempted to evade by means of the alter ego trust. It is the presence of this obligation that constitutes the nexus between the existence of the common law power and it actually being exercised. So, for example, the REM case shows that in the case of a marriage to which the accrual system applies, the obligation sought to be evaded is the obligation to provide a true and accurate reflection of the trustee-spouse’s accrual. In the case of a marriage to which s 7(3) of the Divorce Act applies, the nexus is provided by the possibility of a redistribution order that, if granted, permits a (partial) transfer of assets to a spouse in circumstances in which he or she would otherwise not be entitled thereto because they are married out of community of property. (For suggestions regarding marriages that are in community of property – which is a particularly tricky scenario – or marriages entered into with complete separation of property after the enactment of the MPA, see Smith 2017 at 2 – 7 and 13 – 15).

In summary, the judgment in the REM case is to be lauded for creating certainty in respect of trust assets and marriages involving the accrual system and for correcting the erroneous approach to legal standing that was taken by the same court in the WT case. In addition, the principles enunciated in this judgment are in my view capable of being applied in other matrimonial property regimes in the manner alluded to above. The REM judgment has thus contributed to making piercing a more viable remedy for aggrieved spouses.

In conclusion, it should be noted that there are a number of contentious aspects of the REM judgment that considerations of space do not permit to be discussed here. The reader is referred to BS Smith ‘Perspectives on the juridical basis for taking (the value of) trust assets of alter ego trusts into account for the purposes of accrual claims at divorce: REM v VM’ 2017(4) SALJ forthcoming for an in-depth analysis.

Bradley Smith BCom LLD (UFS) is an Associate Professor of Private Law at the University of the Free State.

This article was first published in De Rebus in 2017 (Aug) DR 22.