11/09/2026
I recently read the judgement in:
Taxpayer A v Commissioner for the South African Revenue Service (VAT 22498) [2026] ZATC 7
Delivered: 27 July 2026
This case dealt with the apportionment of input VAT under section 17(1) of the VAT Act.
What caught my attention was the taxpayer’s request to use a transaction-count method across the business. SARS instead approved a dual approach, using different methods for different types of expenditure.
The Court ultimately dismissed the taxpayer’s appeal, finding that there wasn’t enough evidence to show that the transaction-count method was the most appropriate method for the business as a whole.
What I found particularly interesting is that the Court did not say transaction counting is necessarily wrong. It recognised that it can be appropriate in certain circumstances — for example, where IT costs are closely linked to the number of transactions processed.
My takeaway?
A VAT methodology shouldn't simply be chosen because it produces a better deduction. It needs to reflect the actual economic use of the inputs and be supported by solid evidence.
That distinction matters. 👀