As a result of these limitations, the established practice required a resident trust to make a distribution or advance a loan to a resident individual beneficiary. This individual would then externalise the funds using their annual foreign investment allowance to fund an offshore trust, typically via a loan bearing market-related interest. This approach created severe cash-flow mismatches and commercial friction. The resident individual suffered a “dry” income tax charge on the deemed or actual interest (often under section 7C of the Income Tax Act 58 of 1962 (ITA)), having to fund the tax liability without receiving the underlying cash. Concurrently, the offshore trust, lacking liquidity, saw its capital growth eroded by compounding interest charges causing the offshore trust to become technically insolvent. Furthermore, routing capital through a resident beneficiary brings those assets back into their personal estate, exposing the wealth to future estate duty and defeating the primary objective of the offshore estate planning structure.

In a significant policy shift, the SARB relaxed these provisions, signalling a recognition that legitimate offshore trust structures play a vital role in international estate planning, provided transparency is maintained. Following the relaxation of the exchange control framework, the South African Revenue Service (SARS) introduced a manual tax compliance process whereby taxpayers must obtain a letter of compliance which is then presented to the authorised dealer to facilitate the capital export. To this extent, the administrative oversight shifted from SARB to SARS, necessitating a careful consideration of the stance taken by SARS in respect of this arrangement. 

This approval is, however, subject to certain requirements. The non-resident trust must be a specifically named beneficiary of the resident trust, and the distribution must align with the terms and conditions of the local trust deed. Furthermore, the resident trust must demonstrate tax compliance, proving that all tax liabilities in respect of the distribution have been or will be settled. Once the compliance letter is issued, it must be submitted to an authorised dealer to lodge the capital export application.

While the regulatory mechanics have been clarified, a substantive question remains regarding the income tax and capital gains tax consequences of a direct trust-to-trust distribution. A primary area of debate is whether such a distribution triggers the anti-avoidance attribution rules.

Donations Tax

From a donations tax perspective, the position is relatively clear. No donations tax should be levied on distributions by a resident trust to a beneficiary, as section 56(1)(l) of the ITA specifically excludes property disposed of under and in pursuance of any trust. As confirmed by the Supreme Court of Appeal in Abraham Krok Trust v Commissioner for South African Revenue Service (58/2010) [2010] ZASCA 153; [2011] 2 All SA 591 (SCA); 73 SATC 105 (29 November 2010), the legislature deliberately enacted this section to ensure that all valid trust distributions, executed in accordance with the trustees’ powers and the trust objects, are unequivocally protected from donations tax.

The Section 7(8) Attribution Threshold

The far more contentious issue is, however, whether a trust-to-trust distribution constitutes a “donation, settlement or other disposition” for the purposes of section 7(8) of the ITA. Section 7 contains a series of deeming provisions which act as statutory anti-avoidance measures. As established in Estate Dempers v SIR 1977 (3) SA 410 (A), 39 SATC 95 these provisions are fundamentally aimed at combating schemes where a taxpayer attempts to minimise liability by diverting income-producing property away from themselves. In this context, section 7(8) targets the expatriation of income, stipulating that if any donation, settlement or other disposition by a resident results in income accruing to a non-resident, that income must be attributed back to the resident donor.

Because historic exchange control restrictions prevented direct cross-border trust distributions, there is a distinct lack of judicial precedent addressing section 7(8) in this specific context. Despite this vacuum, SARS appears to have adopted a broad and potentially overreaching interpretation in its Comprehensive Guide to Capital Gains Tax, asserting that the vesting of assets by a trust may constitute a disposition for the purposes of the attribution rules. The critical enquiry is not simply whether the distribution constitutes a donation or settlement, but whether the trustees can properly be said to have made a disposition of the kind contemplated by section 7(8). Section 7 is directed at arrangements whereby a taxpayer voluntarily diverts income-producing property while avoiding the incidence of tax. In this regard it is submitted that a trustee occupies an entirely different legal position.

It is submitted that a distribution of trust capital by a resident trust to an offshore trust does not constitute a donation as understood in the legislation or at common law. A donation requires a gratuitous disposal of property. The courts have consistently held that this necessitates an element of liberality or generosity on the part of the donor, who must act voluntarily while receiving nothing in return. In the context of a discretionary trust, trustees do not possess the capacity to act out of personal liberality. They are firmly constrained by the trust deed and bound by a strict fiduciary obligation to distribute property solely for the benefit of the beneficiaries. Consequently, the distribution by a trustee is the performance of a pre-existing legal duty, completely lacking the gratuitous element essential to a donation. Similarly, the transaction is not a new settlement of property, but rather the administrative implementation of assets already settled into the resident trust estate.

Put differently, trustees possess no beneficial interest in the trust assets, derive no personal economic benefit from their administration and cannot voluntarily impoverish themselves by distributing property. They merely perform fiduciary obligations imposed by the trust deed in administering assets previously settled upon the trust. Properly characterised, the trustees are not making a fresh disposition of their own property but implementing the founder’s original disposition in accordance with the terms of the trust.

Qualifying Dispositions

Having excluded a donation, the analysis turns to the phrase “settlement” or “other disposition”. Applying the ejusdem generis rule, this term must be restricted to dispositions sharing the essential characteristics of a donation, namely an appreciable element of gratuitousness. As the Appellate Division confirmed in Ovenstone v SIR 1980 (2) SA 721 (A), 42 SATC 55 and Joss v SIR 1980 (1) SA 674 (T), 41 SATC 206, these phrases cannot capture every legally recognised alienation of property. It is a misconception to suggest that the mere absence of a commercial quid pro quo elevates a fiduciary distribution to a statutory disposition. The trustees act out of a binding legal mandate, which remains fundamentally distinguishable from the benevolent transfers section 7(8) was designed to capture. It is submitted that the operative disposition occurred when the founder settled assets into the resident trust, not when the trustees later exercise their discretion.

To demonstrate the theoretical hazard of interpreting a trust-to-trust distribution as an attributing disposition, one must consider the practical consequences. Imagine a resident trust founded decades ago that has accumulated capital upon which it has paid all required income taxes. The trustees resolve to vest a portion of this capital in a non-resident beneficiary, who then uses the funds to purchase an investment property in London. If SARS’ broad interpretation holds, the rental income derived from that foreign property would be perpetually attributable to the South African trust. While section 6quat of the ITA does provide a theoretical rebate mechanism for foreign taxes paid by the offshore beneficiary in such attribution scenarios, its practical application creates an administrative nightmare. To claim this rebate, the South African trustees would bear the onus of sourcing final foreign tax assessments from offshore beneficiaries, navigating mismatched foreign tax years, applying complex exchange rate translations and performing intricate proportionate limitation calculations. This broad interpretation effectively forces local trustees to become shadow accountants for non-resident beneficiaries, imposing a perpetual and unworkable compliance burden on the local trustees.

To aggravate matters, paragraph 72 of the Eighth Schedule functions as the capital gains counterpart to section 7(8). Should the broad interpretation be accepted, a resident trust would be transformed from a tax-neutral conduit into a permanent tax-reporting surrogate for the offshore trust. If the offshore trust invests the distributed funds into foreign equities and later realises a profit, that gain would have to be included in the resident trust’s aggregate capital gain. This imposes a perpetual compliance burden on local trustees, forcing them to perform complex deemed-resident calculations on the gains of a foreign entity over which they exercise no administrative control.

The Causal Nexus and the Corporate Veil

Even if a court were to classify a trust-to-trust distribution as a qualifying disposition, the attribution rules are further limited by the requirement of a causal nexus. Section 7(8) only applies to amounts that would have constituted income had the non-resident been a resident. If the offshore trust utilises the distribution to invest in a foreign company that generates dividend income through operating subsidiaries, one must ask if that dividend would constitute taxable income in South Africa? Under section 10B of the ITA, a resident holding at least 10% of the equity shares and voting rights (but less than 50%) in a foreign company is exempt from normal tax on foreign dividends. Consequently, the dividend would not constitute income, limiting the potential for attribution. Furthermore, where the shareholding is 50% or more, the controlled foreign company rules under section 9D apply. This demonstrates that the legislature already has a mechanism for taxing offshore corporate income, thus rendering a broad application of section 7(8) to underlying company profits rather inappropriate.

In addition, SARS would face significant legal hurdles in looking beyond the offshore trust to attribute the trading income of the underlying foreign subsidiaries to the South African trust. The legislation requires the income to be “attributable to” the disposition, necessitating an unbroken causal chain. Where an offshore trust invests funds into an active corporate structure, the proximate cause of the subsequent income is the independent corporate machinery and commercial operations of the foreign entities, not the initial trust distribution. This view aligns with the principles in CIR v Widan 1955 (1) SA 226 (A), 19 SATC 341, where the court acknowledged, albeit under a different statutory context, that intervening factors such as business management and labour attenuate and break the causal link to the original funding. This position is also confirmed by Binding Private Ruling 157, wherein SARS noted that section 7(8) limits attribution to amounts generated directly in the trust and does not extend to amounts generated by underlying corporate entities. Consider another practical example where an offshore trust receives a trust-to-trust distribution and subsequently advances those funds as an interest-bearing loan to its wholly-owned offshore operating company. The company uses the loan to fund its active operations, generating trading profits, while paying interest back to the offshore trust. If the requirement for a strict, proximate causal nexus is ignored, SARS might attempt to look through the corporate veil and attribute not only the interest received by the trust, but also the active trading profits generated by the subsidiary back to the South African trust. This would be legally untenable as it conflates the yield on the funding with the fruits of the business, ignores the separate legal personality of the subsidiary and fails to recognise that the proximate cause of the trading profit is owing to independent commercial activities, the independent corporate machinery, management expertise and use of labour of that company rather than the initial funding in the form of the first trust distribution. Stated differently, it would conflate the contractual return on capital with independent fruits of a separate business enterprise.

Conclusion

In summary, it is submitted that trust-to-trust distributions do not constitute a “donation, settlement or other disposition” as contemplated in section 7(8). Relying on the principles established in Ovenstone, the fiduciary nature of trust distributions wholly distinguishes them from the gratuitous transfers targeted by the anti-avoidance rules. Furthermore, even if the rules were engaged, the requisite causal nexus is effectively severed when distributed assets are pooled and subjected to the independent commercial operations of foreign entities.

Notwithstanding these substantial technical grounds, it is appreciated that taxpayers are inherently risk-averse, and the absolute lack of binding judicial authority warrants a measured approach. It is likely SARS is already considering ruling requests on the subject of trust-to-trust distributions. Given the technical complexity, trustees are advised to closely monitor these administrative developments. Should a ruling be issued that is insufficiently broad or inconsistent with the principles of fiduciary administration, the recommended course of action is to submit a bespoke application for a Binding Private Ruling on behalf of the specific trust where the commercial value of certainty justifies the cost. Such a ruling will provide the required absolute regulatory certainty, ensuring that families can safely utilise the relaxed exchange control landscape to achieve their intergenerational wealth planning objectives without the risk of SARS taking a different view.