SARS has published a draft model and proposed legislation to establish an advance pricing agreement programme and is inviting comments on it by the end of January.
By Karen Miller, consultant, and Joon Chong, partner at Webber Wentzel
In 2020, SARS released a draft discussion paper on the introduction of an advance pricing agreement (APA) programme, which is intended to provide taxpayers with clarity and certainty on their transfer pricing (TP) obligations. Overall, comments from the public were supportive.
SARS has now released a proposed model for establishing the APA programme, as well as draft legislation (proposed model). In preparing this, SARS did take on board comments made on the previous discussion paper. We welcome the publication of this much-awaited draft as it demonstrates SARS’ commitment to provide certainty for taxpayers in a complex and subjective area of taxation. The proposals are also a positive step towards avoiding costly TP disputes for multinationals operating in South Africa.
While this is a welcome development, we are a little sceptical about the timing of the implementation and the process outlined in the proposed model. SARS does not have a good track record in negotiating TP matters under the existing mutual agreement procedure.
SARS intends to put a pilot programme in place as soon as the enabling legislation has been passed. Interestingly, SARS has shied away from considering unilateral APAs[1], indicating it will only consider bilateral APAs[2] as part of its pilot project. The proposed pilot also suggests that SARS will only accept APA applications under its pilot programme with OECD member countries. It hopes to gain valuable experience, learn from other jurisdictions, and expand its capacity before it introduces the full programme or accepts multilateral APA applications. Given this intention, it is likely that the initial bilateral APA applications will be with foreign revenue authorities that have historically demonstrated their willingness to share insights and experience with SARS.
This is disappointing on two fronts. Firstly, a unilateral APA could provide a valuable extension to the existing alternative dispute resolution process to provide taxpayers with some certainty on open years following the resolution of an audit. SARS will also not allow the outcome of an agreed APA to be rolled back, i.e., to have the negotiated position in the APA applied to similar historical transactions as part of an audit resolution. This is unfortunate, as both the unilateral APA and rollback are likely to be more efficient to negotiate and can be used in settlement negotiations, providing much-needed prospective clarity and finalization of costly TP disputes for taxpayers.
Secondly, many South African-based multinationals experience TP audits into transactions with other African jurisdictions. Resolution of these is often one-sided, resulting in significant unrelievable double taxation.
The proposed model observes that the main impediment in establishing a workable APA programme is the lack of TP expertise in South Africa. Negotiations on an APA should be undertaken by a team independent of the TP audit teams. However, due to this lack of TP expertise, there will be an exchange of expertise and staff between the two teams in the early stages.
This creates some concern, as it will be essential for SARS to establish clear boundaries between the APA unit and the TP unit to prevent information shared during the APA process from being used in a TP audit, notably if the initial application by the taxpayer is rejected for an APA by SARS. SARS personnel involved in an APA process should not be involved in a TP audit on the transaction which is the subject of the APA. Access to information disclosed during an APA process should be restricted and available only to the SARS officials involved in the decision-making relating to the APA application process. If the APA process is unsuccessful, information shared during the process should not be used to audit the transaction or multinational group. As SARS has indicated it will require substantive information to be supplied in the pre-application process, establishing these boundaries could create a challenge for SARS. We therefore hope that SARS will implement the correct measures to ensure such information will not be shared with the rest of SARS, including with the other members of the TP unit.
An APA will consider a specific transaction and the most appropriate TP method to test that the pricing (or resulting profit achieved) is arm’s length. It provides certainty of relief from double taxation by ensuring the appropriate allocation of profit and, if necessary, any compensating adjustments to ensure no double taxation is suffered.
The APA will contain, among other things, “critical assumptions” that should apply to the affected transactions. We hope that SARS will provide more guidance on the scope of what they would accept as critical assumptions. Examples of critical assumptions include assumptions of economic conditions, functions and risks of enterprises involved in the affected transaction, or interpretations of South African and foreign tax law.
The draft legislation provides for taxpayers who have entered into APAs to submit “compliance reports” to SARS within 60 days of the end of the tax period within the duration of the APA, or from the date of termination of the APA. These reports should confirm that, among other things, circumstances in the relevant year are unchanged from those in the APA application. They should demonstrate compliance with the terms and conditions of the APA, and confirm that the critical assumptions in the APA have been complied with. It is uncertain whether these compliance reports would also be provided to the foreign tax authority and what would happen if the foreign tax authority disagreed on whether the critical assumptions have been complied with. This begs the question as to how successful the programme will be, especially if and when it is extended to all South Africa’s treaty partners.
The draft legislation also enables SARS to terminate an APA retrospectively and sets out various instances when this could occur, for example, if the effect of the APA “will materially erode the tax base and it is in the public interest to withdraw or modify the APA retrospectively” or there is a breach of critical assumptions.
SARS can terminate an APA if a court overturns or modifies an interpretation of legislation on which the agreement is based. In that case, the agreement will cease to be effective from the date of judgement unless (i) the decision is under appeal; (ii) the decision is fact-specific and general interpretation upon which the APA was based is unaffected; or (iii) the reference to the interpretation upon which the APA was based was obiter dicta. It will be interesting to see how an APA can be impacted by future decisions as affected transactions are usually very fact-specific.
Webber Wentzel will be providing submissions on the proposed model and draft legislation to SARS by the due date, which is 31 January 2022.
[1] An APA entered into between the taxpayer and SARS only – akin to a binding ruling
[2] An APA entered into between two tax jurisdictions for two taxpayers