Taxpayer EPP v CSARS, VAT, section 23C and the route back to the correct year

The Tax Court’s judgment in SARSTC IT 24852 (IT) ZATC JHB (14 April 2026) is, at one level, a fairly orthodox section 11(a) timing case.

The taxpayer claimed a deduction in the wrong year. SARS refused it. The Tax Court agreed with SARS. Appeal dismissed.

But that is not the interesting part.

The interesting part is what happens next.

Because if SARS and the Tax Court are right that the expenditure was not deductible in 2015 because it was incurred in the 2011 to 2013 years of assessment, then the immediate practical question is obvious:

Can the taxpayer now go back to the 2011 to 2013 years and fix the assessments for those years?

That question is not a footnote to the judgment. It is the whole procedural fight that now follows.

The background

The taxpayer was a licensed distributor of fuel. It purchased fuel from South African manufacturers and resold that fuel to customers outside South Africa.

When it purchased the fuel, the taxpayer paid excise duties and levies in addition to the purchase price.

Because the fuel was exported, the taxpayer was entitled, subject to compliance with the Customs and Excise Act, to claim a refund of those duties and levies. The refund claims had to be submitted within the relevant statutory time period.

That did not happen.

The taxpayer became aware during 2013 that its clearing agent had not prepared and submitted certain refund claims timeously. The refund claims relating to the 2011 to 2013 tax periods had prescribed.

The taxpayer did not claim the excise duties as deductions in the relevant earlier years. It also did not include them in trading stock costs for purposes of section 22 of the Income Tax Act.

Instead, in its 2015 income tax return, the taxpayer claimed R38 831 547 as a deduction under section 11(a) of the Income Tax Act. SARS disallowed the deduction, imposed a 10% understatement penalty and levied interest under section 89quat.

The matter went to the Tax Court.

The issue

The main issue was whether the taxpayer was entitled to deduct the R38 831 547 in 2015.

The taxpayer’s case was not simply that it had paid the duties in 2015. It had not.

Its case was more sophisticated than that.

The taxpayer argued that it was not claiming old expenditure in 2015. It was claiming a loss.

On this argument, the taxpayer still had a refund right until the Customs and Excise Act time period expired. The loss was therefore only incurred when that refund right became irrecoverable. The loss, so the argument went, was involuntary and fortuitous. It was caused by the clearing agent’s failure to lodge the refund claims timeously.

The taxpayer therefore distinguished between “expenditure” and “loss” for purposes of section 11(a). Expenditure is generally a voluntary outlay. A loss is generally an involuntary deprivation. The taxpayer said this was the latter.

SARS disagreed.

SARS’ case was that the taxpayer was trying to shift expenditure from the years in which it was actually incurred to a later year by calling it a loss.

According to SARS, the taxpayer incurred the expenditure when it purchased the fuel and paid the excise duties and levies. That happened in the 2011 to 2013 years. The later failure to claim a refund under the Customs and Excise Act did not convert the earlier expenditure into a new loss in 2015.

The judgment

The Tax Court agreed with SARS.

The court held that the true nature of the amount claimed was not a new loss incurred in 2015. The excise duties and levies were paid as part of the taxpayer’s fuel purchases. They were incurred in the ordinary course of the taxpayer’s trading operations. They were not fortuitous or involuntary in the relevant sense.

The court accepted that the refund entitlement was a separate statutory mechanism. The taxpayer’s failure to recover the refund did not change the character or timing of the original outlay.

The court also relied on the annual nature of income tax. Expenditure incurred in one year cannot generally be moved to another year because it is more convenient to claim it there. If the duties were deductible, they were deductible in the years in which they were incurred.

The taxpayer’s appeal against the 2015 assessment was therefore dismissed.

The Tax Court also upheld the understatement penalty. The taxpayer had deliberately adopted a tax position. The court was not dealing with a mere bona fide inadvertent error in completing a return. The prejudice to SARS exceeded R1 million and the 10% penalty was upheld.

Interest was also left in place. The court held that the circumstances giving rise to the interest were not beyond the taxpayer’s control.

The VAT knee-jerk reaction

The immediate response from many tax practitioners will probably be this:

Is this not just the same as a VAT input tax claim that prescribes?

A vendor receives an invoice. The vendor pays the full amount, including VAT. The VAT component is not deducted for income tax purposes because it is claimable as input tax. If the vendor later fails to claim the input tax before the VAT prescription period expires, should the VAT then become deductible for income tax purposes?

At first blush, this analogy is attractive.

But it is not complete.

The reason input tax is generally not deducted for income tax purposes is not that the VAT component was never “actually incurred”. As between the vendor and the supplier, the full invoice liability, including VAT, is incurred.

The reason the input tax component is excluded is section 23C of the Income Tax Act.

Section 23C provides, in essence, that where regard must be had to the cost of an asset or the amount of expenditure incurred, and the taxpayer is a VAT vendor who is or was entitled under section 16(3) of the VAT Act to deduct input tax, the input tax must be excluded from the cost, market value or expenditure.

The words “is or was” matter.

If the vendor was entitled to deduct input tax, section 23C does not easily disappear merely because the vendor failed to claim the input tax in time. That is a statutory answer. It is not a metaphysical debate about whether VAT was “actually incurred”.

That makes VAT different from the excise duty issue in this case.

There was no equivalent of section 23C for the customs and excise refund entitlement. There was no income tax provision saying that if the taxpayer is or was entitled to a refund under the Customs and Excise Act, the amount of that potential refund must be excluded from expenditure.

The Tax Court therefore had to deal with the matter under ordinary section 11(a) principles: what was the amount, what was its character, and in which year was it incurred?

That is why the VAT analogy is useful as a conversation starter, but dangerous as an answer.

VAT is different because section 23C says it is different.

The more important issue: what now?

The taxpayer lost 2015.

But the judgment does not simply say: no deduction.

It says: no deduction in 2015 because the expenditure was incurred in 2011 to 2013.

That matters.

Because if the expenditure was incurred in 2011 to 2013, then one has to ask why the taxpayer should not now request reduced assessments for those years.

This is where the judgment becomes less about section 11(a) and more about the Tax Administration Act.

The first takeaway: get the year right

The first lesson is obvious but brutal.

The correct year of assessment is not a technicality.

It can decide the whole case.

This judgment is a reminder that a taxpayer must identify the correct year in which the deduction arises before deciding on the procedural route. If the taxpayer challenges only one year, and that year turns out to be wrong, the taxpayer may win the legal theory but lose the assessment.

That seems to be what happened here.

The taxpayer challenged the 2015 assessment. It seemingly did so because it was convinced that the deduction arose in 2015 when the refund claims prescribed. That was a coherent argument, but it was also an all-or-nothing argument.

If the taxpayer was right, 2015 was the correct year.

If the taxpayer was wrong, 2015 was not merely a weak year. It was the wrong year.

There may also have been a procedural reason for focusing on 2015. Perhaps the earlier years were already prescribed. Perhaps there was a concern that SARS would refuse to reopen them. Perhaps the taxpayer thought it was procedurally cleaner to fight the 2015 assessment and then deal with the earlier years later.

Whatever the reason, the result is the same: the taxpayer now has a judgment saying the 2015 year is wrong and the earlier years are the years that matter.

That creates the next question.

Can the taxpayer use the judgment to open the earlier years?

Timing is not just timing

This is also the broader lesson for accountants, auditors and tax practitioners.

Timing issues are often underplayed because they are seen as cash-flow issues. The thinking is familiar: if the deduction is wrong in one year, it will reverse in another year; therefore, there is no real loss to the fiscus.

That is a dangerous shorthand.

A wrong-year deduction can create two separate risks.

First, the deduction may be lost entirely if the correct year cannot be reopened. That is not a timing difference anymore. It is a permanent disallowance caused by a procedural problem.

Secondly, the wrong-year claim can still create an understatement penalty exposure. The fact that the dispute feels like timing does not mean that the taxpayer is immune from an understatement penalty. If the deduction reduced tax in the wrong year, SARS may still say there was prejudice in that year.

Wear-and-tear claims are a simple example. A taxpayer may claim an allowance too early. Over the life of the asset, the allowance may broadly even out. But in the year under assessment, SARS may still say the taxpayer claimed too much too soon. If the correct later year is not open, the taxpayer may lose the allowance. If SARS imposes a USP, the taxpayer must still answer the penalty case.

That is why the year of assessment is not just an accounting allocation. It is a substantive risk point.

The second takeaway: the post-judgment move is a reduced assessment request

The post-judgment move is not, in my view, to treat the judgment as the end of the road.

The post-judgment move is to file reduced assessment requests for the 2011 to 2013 years.

The request would be brought under section 93(1)(d) of the TAA, which allows SARS to issue a reduced assessment if SARS is satisfied that there is a readily apparent undisputed error in the assessment by SARS or by the taxpayer in a return.

Section 93(2) is also important. It provides that SARS may reduce an assessment even though no objection has been lodged or appeal noted.

So the starting point is this: the fact that the taxpayer did not object to the 2011 to 2013 assessments does not, by itself, prevent SARS from issuing reduced assessments.

The difficulty, of course, is prescription.

If the earlier assessments have not prescribed, the route is simpler. The taxpayer asks SARS to correct the earlier assessments under section 93(1)(d) on the basis that the omission of the duties from those years was an error in the returns or assessments.

If the earlier assessments have prescribed, the taxpayer needs an exception to prescription under section 99(2).

There are two possible routes.

Route 1: section 93(1)(a), section 99(2)(d)(i) and I-CAT

Section 99(2)(d)(i) says that the ordinary prescription limits do not apply to the extent necessary to give effect to the resolution of a dispute under Chapter 9 of the TAA.

The reduced assessment mechanism then becomes the way SARS gives practical effect to that resolved dispute.

The argument may be framed through section 93(1)(a): SARS may make a reduced assessment if the taxpayer successfully disputed the assessment under Chapter 9. Depending on the precise procedural posture, section 93(1)(c) may also be relevant because it deals with a reduced assessment necessary to give effect to a judgment pursuant to a Chapter 9 appeal.

This is where the technical debate starts. The taxpayer did not run a separate Chapter 9 dispute for the 2011 to 2013 assessments. But the answer, on the I-CAT logic, is that it may not need to. The reduced assessment for the earlier years is not being used to start a new dispute. It is being used to give effect to the dispute that has already been resolved.

That is exactly the concern I raised in I-CAT. If the resolution of one year necessarily determines what must happen in another year, the correction of the other year may be presented as implementation, not a fresh objection and appeal.

This brings us to I-CAT.

In my previous article on I-CAT, I said that the result may have been correct, but the reasoning was, with respect, problematic. The taxpayer in I-CAT had a Chapter 9 dispute in one year and sought to correct another year. The High Court accepted, on my reading, that the other year could be corrected post-prescription because doing so gave effect to the resolution of the Chapter 9 dispute.

At the time, I warned that this reasoning opens a door.

The door is this: where the resolution of a dispute in one year necessarily determines what must happen in another year, taxpayers may try to argue that the other year is protected from prescription because correcting it gives effect to the resolution of the dispute.

That is exactly the type of argument this taxpayer may now want to run.

The 2015 dispute was a Chapter 9 dispute. The Tax Court resolved it. The court’s resolution was that the deduction did not belong in 2015 because the expenditure was incurred in 2011 to 2013.

So, the taxpayer may argue, correcting 2011 to 2013 is necessary to give effect to the resolution of the 2015 dispute.

Is that argument free from difficulty?

No.

SARS may say that the 2015 dispute has already been fully resolved. The 2015 appeal was dismissed. The 2015 assessment was confirmed. Nothing more is required to give effect to that judgment.

SARS may also say that there was no separate Chapter 9 dispute in relation to the 2011 to 2013 assessments. Those years were not before the Tax Court. The Tax Court did not order SARS to reduce those assessments. The taxpayer’s answer is the I-CAT answer: if the Chapter 9 dispute has already resolved the timing issue, the reduced assessments for the earlier years may be the mechanism to give effect to that resolution, not a second Chapter 9 process for those years.

That is the same type of concern I raised in the I-CAT article. A Chapter 9 dispute is not imaginary. It exists because an objection and appeal process was actually instituted in relation to an assessment.

But I-CAT makes the argument available.

Indeed, this may be the very case where the I-CAT logic is used in the way I warned it could be used. The earlier years were not the years under appeal, but the resolution of the 2015 appeal may arguably include a finding that the amount belongs in those earlier years.

That is not a bad argument. It is also not a safe argument.

It is a litigation argument.

Route 2: section 99(2)(d)(iii) and section 93(1)(d)

The cleaner route, if the facts support it, is section 99(2)(d)(iii).

This is the route I discussed in the L’Avenir series, particularly in Part 2 and Part 3, with the broader procedural warning in Part 4.

Section 99(2)(d)(iii) says that the ordinary prescription periods do not apply to the extent necessary to give effect to an assessment referred to in section 93(1)(d), if SARS became aware of the error before the expiry of the ordinary prescription period.

This is important.

The provision does not say that the taxpayer must have filed a formal reduced assessment request before prescription.

It says SARS must have become aware of the error before prescription.

That is a different thing.

But, it is not enough to say that the Tax Court found against the taxpayer in 2015.

SARS may argue that the Tax Court decided only that the deduction was not allowable in 2015. It did not decide that the amount must be allowed in 2011 to 2013. SARS may also say that there remain questions about section 22 trading stock, quantum, proof of payment, whether the duties were included elsewhere, whether the taxpayer has already received any economic benefit, and whether all section 11(a) requirements are met in the earlier years.

Those are the obvious SARS answers.

The taxpayer’s answer would be that SARS itself ran the case on the basis that the expenditure was incurred in 2011 to 2013. The Tax Court accepted that timing conclusion. The taxpayer’s point would be that the basic timing correction is the very consequence of SARS’ successful argument.

The irony

There is an irony here.

The taxpayer lost because the court accepted SARS’ timing argument.

But that same timing argument may now become the taxpayer’s strongest weapon.

How does SARS say, on the one hand, that the expenditure was incurred in 2011 to 2013 and, on the other hand, refuse to correct 2011 to 2013 because the taxpayer should have claimed the amount in those years?

Of course, SARS may say prescription. In fact that is exactly what SARS said in the I-CAT case.

But prescription is not always the end of the enquiry. Section 99(2) exists precisely because there are circumstances where the ordinary prescription limitation does not apply.

The real question is which exception applies.

If one follows I-CAT broadly, the taxpayer may argue that correcting the earlier years gives effect to the resolution of the 2015 Chapter 9 dispute.

If one follows the cleaner section 99(2)(d)(iii) route, the taxpayer must prove that SARS became aware of the section 93(1)(d) error before prescription.

Both routes require careful factual work.

Neither route should be dismissed out of hand.

Conclusion

This judgment is a warning about year-of-assessment risk and the danger of treating timing as harmless. The often-cited excuse ‘it is just a timing risk’ underestimates the risk. A wrong-year deduction may be lost entirely if the correct year cannot be reopened. It may also create an understatement penalty exposure, even where the dispute looks like timing rather than permanent loss to the fiscus.

The taxpayer in this case may have had a real deduction. The problem was that it claimed the deduction in the year in which the refund claims prescribed, rather than in the years in which the duties were actually incurred.

The Tax Court has now answered the 2015 question. The answer is no.

But the judgment may also have created the foundation for the next question.

If the deduction belongs in 2011 to 2013, can those years now be corrected?

That is where section 93(1)(d), section 99(2)(d)(i), section 99(2)(d)(iii), I-CAT and the procedural lessons from L’Avenir become important.

The case may have been lost in 2015.

That does not automatically mean the deduction was lost forever.