Decision Alert: Federal Court finds PepsiCo liable for royalty withholding tax and potential diverted profits tax

By Ryan Leslie, Geraldine Chan and Graeme Cooper


The Federal Court has ruled in favour of the Commissioner of Taxation (Commissioner) against PepsiCo, Inc (PepsiCo) in PepsiCo, Inc v Commissioner of Taxation [2023] FCA 1490 (30 November 2023). The case concerned PepsiCo’s “royalty-free” agreements with Schweppes Australia Pty Ltd (SAPL).

The Federal Court ruled that PepsiCo was liable for royalty withholding tax (at a 5% rate) in relation to portions of payments made under exclusive botting agreements (EBAs) which were held to be royalties and, in the alternative, diverted profits tax (DPT) would apply at a rate of 40%, i.e. if the payments were not royalties.

This decision was the first time the DPT provisions were considered by a Court. PepsiCo have now appealed the decision. In order to achieve a successful outcome, PepsiCo needs to succeed on both the royalty issue and the DPT issue. If PepsiCo succeeds on the royalty issue but doesn’t also succeed on the DPT issue on appeal, this would increase the tax bill 8-fold.

Key takeaways

  • Scrutiny of “royalty-free” arrangements: the case further emphasises the ATO’s ongoing focus on cross-border intangible arrangements where royalty withholding tax has not been paid, including where licences are granted and no express royalty is payable. Multinational entities operating in Australia under similar arrangements may face increased scrutiny. The ATO has also released a draft taxation ruling TR 2024/D1 Income tax: royalties – character of payments in respect of software and intellectual property rights, which reflects its current draft views on when payments made in respect of software and intellectual property rights will be a royalty. Those views continue to differ from those expressed in the now withdrawn ruling TR 93/12.
  • A broader approach to characterisation of “royalties”?: previous case law in IBM and Task Technology had emphasised that in determining whether payments made are characterised as royalties, Australian courts have interpreted the rights and obligations under the relevant agreement,’ as the agreement will determine whether the payments are for the right or use of, relevantly, software. However, the decision of the Federal Court indicates that in considering whether a payment constitutes a “royalty”, it is necessary to consider the characterisation of the relevant payments and the terms of the relevant agreements in their “business and commercial context”. The Court drew support from section 6(1) of the Income Tax Assessment Act 1936 (Cth) (ITAA 1936) which defined royalty regardless of how the payments are described and that Article 12(4) referred to payments “of any kind”.
  • Payments from an Australian resident entity to another Australian resident entity may give rise to royalty withholding tax: the Federal Court found that the payments made by SAPL were “income derived by” and “paid at the direction of” PepsiCo since the EBA obliged SAPL to buy concentrate from PepsiCo or a subsidiary nominated by PepsiCo. Such a broad reading of the EBA would appear to impose a withholding obligation on SAPL (despite SAPL paying an Australian entity and not being pursued for failure to withhold penalties based on the facts described in the case).
  • DPT: the Federal Court’s decision sheds some light on the application of the DPT provisions. It confirms that the “principal purpose test” within the DPT provisions is a lower bar compared to the “dominant purpose test” within the general anti-avoidance provisions. The Federal Court confirmed that the meaning of “principal purpose” is “a purpose that is a prominent, leading or main purpose”, of which there can be more than one but need not be the dominant purpose.


PepsiCo and Stokely-Van Camp, Inc (SVC), had EBAs with an independent Australian bottling company, Schweppes Australia Pty Ltd (SAPL). Under this agreement, PepsiCo or SVC agreed to sell, or cause a related entity to sell, beverage concentrate to SAPL for bottling and sale, and granted SAPL the right to use the Pepsi and Mountain Dew trademarks. However, no royalties were paid under the agreement, and therefore, no withholding tax was paid in Australia. The only relevant payments made were for the purchase of concentrate. During the 2017-18 and 2018-19 years (the Relevant Years):

  • Concentrate Manufacturing (Singapore) Pte Ltd (CMSPL), a Singaporean entity and a member of the PepsiCo Group, produced concentrate according to a recipe or formula provided by PepsiCo and SVC;
  • CMSPL supplied the concentrate to PepsiCo Beverage Singapore Pty Ltd (PBS), an Australian entity also a member of the PepsiCo Group;
  • PBS supplied concentrate to SAPL and invoiced SAPL for the concentrate that had been supplied and SAPL paid PBS for the concentrate; and
  • PBS transferred the money received from SAPL to CMSPL (we assume as a concentrate purchase price – the ATO did not seem to assert that any portion of this payment was a royalty), retaining only a small margin.

It is interesting to note that the terms of the EBAs were changed in 2020, and the payment and pricing terms were removed from the EBAs and instead moved to a separate Concentrate Sale Agreement. Such a change may have led to a different tax outcome for Pepsico and SVC, however this was not considered by the Federal Court as the change to the terms of the EBAs occurred after the Relevant Years.

The Commissioner issued royalty withholding tax notices under section 128B of the ITAA 1936 to PepsiCo and SVC requiring both entities to pay approximately $3.6 million in royalty withholding tax on money paid by SAPL to PBS.

In the alternative, the Commissioner issued DPT assessments pursuant to Part IVA of the ITAA 1936 imposing a tax liability of $28.9 million on PepsiCo and SVC.


Royalty withholding tax

The Federal Court held that royalty withholding tax was payable because:

  • a portion of the payments made by SAPL to PBS were “consideration for” the use of, or the right to use, the relevant trademarks and other intellectual property which were provided under the EBAs, in the circumstances of this case. As such, the payments satisfied the definition of “royalty” in Article 12(1) of the US DTA and section 6(1) of the ITAA 1936. Although PBS was not a party to the EBAs, the Federal Court held that payments made by SAPL and PBS were linked to the license of PepsiCo’s trademarks and intellectual property (that is, without PepsiCo’s trademarks and IP: (a) SAPL would not have been able to package and sell the beverages under PepsiCo’s and SVC’s brands and if so, it could be inferred that SAPL may not have agreed to make the payments, and (b) a failure by SAPL to perform its payment obligations could result in a termination of the agreement and therefore the licence). The Federal Court’s consideration of the treaty was on a ‘textual’ basis and did not consider extrinsic materials that may have assisted in interpreting it;
  • the circumstances in which a royalty will be found to be payable extend beyond the terms of the agreement, and may include the business and commercial context in which the agreement was struck. In the circumstances, his Honour considered that it was appropriate to look beyond the description of the payments in the agreement between the parties. His Honour concluded that “the licence of the trademarks and other intellectual property was fundamental to the agreement” and that “the payments…were linked with the licence of the trademarks and other intellectual property”;
  • the relevant portions of the payments were income derived by PepsiCo or SVC for the purposes of section 128B(2B)(a) of the ITAA 1936 and amounts to which they were beneficially entitled for the purposes of Article 12 of the US DTA. As PBS was nominated as the seller of the concentrate under the EBAs, the Federal Court held that this constituted a direction to pay PBS rather than PepsiCo/SVC. The payments therefore “came home” to PepsiCo/SVC by being applied as they directed; and
  • for the same reasons above, the relevant portions of the payments were therefore deemed to have been paid by SAPL to PepsiCo or SVC by virtue of section 128A(2) of the ITAA 1936. The payments were “dealt with on behalf of PepsiCo/SVC” or dealt with “as PepsiCo/SVC directs” as the Federal Court concluded that PepsiCo and SVC were entitled to receive the payments made by SAPL under the EBAs and had nominated PBS as the seller of the concentrate under the EBAs, and thereby directed SAPL to pay PBS.

The conclusion that the payments were “income derived by” entities other than the entities which sold the concentrate seems to rely on a very broad interpretation of the relevant concepts.


The Federal Court held that if the royalty withholding tax provisions did not apply, the DPT provisions would apply.

The alleged scheme was entry into the EBAs on terms whereby no royalty was paid for the use of intellectual property. In determining that each of PepsiCo and SVC obtained a “tax benefit”, The Federal Court accepted the Commissioner’s counterfactual that absent the scheme, the EBAs might reasonably have been expected to express the payments to SAPL to be for all of the property provided by PepsiCo rather than for concentrate only. Consequently, PepsiCo/SVC might reasonably be expected to have been liable to pay royalty withholding tax on a portion of the payments.

The Federal Court ruled that:

  • each of PepsiCo and SVC obtained a tax benefit in connection with the relevant scheme; and
  • having regard to the matters in section 177J(2) of the ITAA 1936, it would be concluded that PepsiCo and SVC in entering into or carrying out the relevant scheme had a principal purpose to obtain a tax benefit (not being liable to pay Australian royalty withholding tax) and to reduce foreign tax (US tax on their income). Relevantly, the Federal Court considered that one of the factors indicating the presence of the requisite principal purpose was the disconnect between the form and substance of the EBAs. In form, the payments to be made by SAPL were for the concentrate alone and not for the licence of the trademarks and other intellectual property. However, in substance, the payments to be made by SAPL were for both the concentrate and the licence of the trademarks and other intellectual property.

Although PepsiCo was unsuccessful in arguing that the royalty withholding tax provisions did not apply to the arrangement, the tax liability for PepsiCo and SVC would have been significantly higher if the royalty withholding tax provisions did not apply and the DPT applied instead (5% of the royalty amount under the royalty withholding tax provisions compared to 40% of the royalty amount under the DPT provisions).

In view of PepsiCo’s decision to appeal the decision, we expect that a superior court’s view on the issues will be instructive on both the royalty and DPT issues, though it of course remains possible a settlement is reached before an appeal is heard given the downside risk if the royalty withholding tax aspects of the appeal succeed and the DPT aspects do not.

Key contacts

Ryan Leslie
Ryan Leslie
Partner, Melbourne
+61 418 186 499
Geraldine Chan
Geraldine Chan
Solicitor, Melbourne
+61 3 9277 2706

Graeme Cooper
Graeme Cooper
Consultant, Sydney
+61 2 9322 4081

Tax Insight: Signs of BEPS Life

The Inclusive Framework of the OECD/G20 has released an “Outcome Statement” on developments in the long-running BEPS project. The Statement is evidence of life for the BEPS project. It focuses on work done but not yet public, rather than announcing significant progress on achieving milestones, new developments or major departures from what was already in the public domain. But, importantly, it also hints at problems in finalising the BEPS project.

1.     Some of the lead up

July 2021. After some false starts in earlier documents, the principal design features of the “Two Pillar” BEPS plan had been largely settled by mid-2021.

The Statement on a Two-Pillar Solution to Address the Tax Challenges Arising from the Digitalisation of the Economy (1 July 2021) laid out the key design features of the 2-pillar BEPS plan:

  • Pillar One would comprise of two elements:
    • Amount A: a portion (between 20-30%) of the surplus profits (profitability above 10% of revenue) of MNEs with global turnover above € 20 bn would be taxable in each market jurisdiction where the MNE earns at least €1 m in revenue (or €250,000 for smaller countries) based on the amount of revenue earned in the jurisdiction;
    • Amount B: calculating the profit from baseline marketing and distribution activities occurring in a country would be simplified and streamlined;
    • Digital Services Taxes: part of the package for developed countries agreeing to Amount A is that countries would agree remove all Digital Service Taxes and similar taxes;
  • Pillar 2 would involve three elements:
    • Income Inclusion Rule: a parent entity would be required to pay top-up tax on the profits of low-taxed members of the group (effective tax rate <15%);
    • Undertaxed Payment Rule: other members of the group would be denied deductions sufficient to collect the top-up tax on the low-taxed income of other group members if the IIR had not been enlivened;
    • Subject to Tax Rule: a treaty-override rule allowing source jurisdictions to impose additional tax (between 7.5% and 9%) on certain classes of payments being made to related low-taxed entities.

The original timetable envisioned Amount A of Pillar One and Pillar Two would both become operative during 2023.

October 2021. The subsequent Statement on a Two-Pillar Solution to Address the Tax Challenges Arising from the Digitalisation of the Economy (8 October 2021) did not fundamentally change that design although it did flesh out some details and tweak a few others:

  • the share of revenue available to market jurisdictions under Amount A of Pillar One was fixed at 25% of residual profit;
  • the amount of low-taxed income immune from being taxed under the IIR and UTPR was increased;
  • a de minimis exemption from the UTPR would operate for the first 5 years for MNEs that have modest tangible assets abroad and operate in fewer than 5 other jurisdictions;
  • source jurisdictions would be allowed to collect up to 9% on payments caught by the Subject to Tax Rule;
  • as well the existing commitment to repeal existing DSTs, countries also agreed not to enforce any new DSTs enacted after 8 October until 31 December 2023 (or the coming into force of the expected Multilateral Convention). This latter commitment would become important in 2023.

December 2021 to March 2022. It was not until December 2022 that a major change to the design emerged with the release of three technical documents on Pillar Two: the Model Rules (20 December 2021), the Commentary (14 March 2022) and Illustrative Examples (14 March 2022).

The Model Rules contained provisions recognising a “Qualified Domestic Minimum Top-up Tax”: a domestic tax on the low-taxed profits of members of a MNE group, collecting any additional tax for the benefit of the source country rather than the country of the parent (under the IIR) or other countries in which the MNE group operates (under the UTPR). This was a significant development, effectively supplanting the IIR and UTPR as the main means of ensuring the global minimum 15% tax rate.

Australian developments. In Australia, the Government has been working toward implementing the basic design:

  • the ALP announced its policy for the May 2022 election which included a promise to implement “a global 15 per cent minimum tax [Pillar Two], and ensuring some of the profits of the largest multinationals – particularly digital firms – are taxed where the products or services are sold [Pillar One]” (22 April 2022);
  • Treasury released a consultation paper, Global agreement on corporate taxation: Addressing the tax challenges arising from the digitalisation of the economy (4 October 2022) seeking submissions on implementation issues; and
  • in the 2023 Budget (9 May 2023), the Government announced the timetable for implementing Pillar Two in Australian law:
    • the Income Inclusion Rule would apply for income years starting on or after 1 January 2024;
    • the Undertaxed Profits Rule would apply for income years starting on or after 1 January 2025; and
    • the 15% domestic minimum top-up tax would apply for income years starting on or after 1 January 2024.

2.     The Outcome Statement

The Inclusive Framework has released its Outcome Statement on the Two-Pillar Solution to Address the Tax Challenges Arising from the Digitalisation of the Economy (11 July 2023) reporting signs of BEPS life in anticipation of the meeting of G20 Finance Ministers and Central Bank Governors’ meeting in India later this month. The Statement was agreed to by 138 members of the Inclusive Framework. There are currently 143 members; Canada was a notable absentee for reasons discussed below. Yes, Canada!

The Outcome Statement is largely a progress report:

  • Pillar One Amount A: the Multilateral Convention allowing countries to claim in domestic law tax on a portion of the residual profits of in-scope MNEs has been prepared for signature: the Statement says “the Inclusive Framework has delivered a text of a Multilateral Convention …” But the Multilateral Convention has not been signed, nor even opened for signing because “a few jurisdictions have expressed concerns with some specific items in the Multilateral Convention. Efforts to resolve these issues are underway with a view to prepare the Multilateral Convention for signature expeditiously.” Just what the text says remains secret as it has not been publicly released (although a small extract, dealing with removal of digital services taxes, was released in December 2022). The new timetable is that the convention should be open for signature in the second half of 2023 “with the objective of enabling the Multilateral Convention to enter into force in 2025”;
  • Pillar One Amount B: the Outcome Statement reports no new developments saying, “further work will be undertaken” on various aspects of the proposal “to be completed by year end” leading to new “content … which will be incorporated into the OECD Transfer Pricing Guidelines by January 2024”;
  • Pillar One Digital Services Taxes: the stand-still agreement on enforcing Digital Services Taxes introduced after October 2021 was conditionally extended from 31 December 2023 to 31 December 2024 (and possibly 31 December 2025). Interestingly, the agreement to delay implementation is conditional on “at least 30 jurisdictions accounting for at least 60 percent of the Ultimate Parent Entities … of in-scope MNEs signing the MLC before the end of 2023”; and
  • Pillar Two Subject to Tax Rule: the drafters have apparently decided that a two-pronged attack is needed:
    • they have “completed and delivered” a model provision (and commentary) to add the STTR to existing bilateral treaties; and
    • they have also prepared a Multilateral Instrument and Explanatory Statement (presumably along the lines of the 2018 Multilateral Instrument) allowing signatories to amend existing bilateral tax treaties in bulk. The new Multilateral Instrument will be open for signature from 2 October 2023.

Neither text was released for scrutiny.

3.     Some observations

It is notable that the Outcome Statement focuses on the BEPS topics of most concern to developing countries, notably Amounts A and B of Pillar One and the Subject to Tax Rule element of Pillar Two: the Statement refers to special provisions which will be in the MLC “designed to address the unique circumstances of developing Inclusive Framework members”, how advancing Amount B “is a critical component of the broader agreement on Pillar One” and how “the STTR is an integral part of achieving consensus on Pillar Two for developing Inclusive Framework members.”

But the Statement hints at problems in finalising several key components of the BEPS project, especially with regard to Pillar One.

The first sign of trouble is with the 2021 compromise – to defer implementing new DSTs until 31 December 2023, in the expectation that the Multilateral Convention would be in place. The hold-outs with regard to the Multilateral Convention (the “few jurisdictions [which] have expressed concerns with some specific items in the MLC…”) are prolonging delivery of the Multilateral Convention. The majority of the members of the Inclusive Framework were willing to prolong the negotiations and extend the moratorium on implementing new DSTs for a further year, but this was too much for Canada. Its DST is scheduled to start on 1 January 2024 in accordance with the original timetable. The Finance Minister was unwilling to accept further delays noting there was still no “firm and binding multilateral timeline to implement Pillar One” and without that, “Canada [was] at a disadvantage relative to countries which have continued collecting revenue under their pre-existing DSTs”. (It is not clear whether the hold-outs are developing countries, who were expecting Amount A to deliver more revenue to them, or the developed countries since their companies are the most affected).

No doubt, Canada is right to be skeptical, but there must be a high chance that the extension of the moratorium will not actually happen because of the conditions attached to the extension. The extension is conditional on “at least 30 jurisdictions … signing the MLC before the end of 2023 …” and those 30 jurisdictions must represent “at least 60 percent of the Ultimate Parent Entities (UPEs) of in-scope MNEs …” Some researchers have concluded this second condition means the US must sign the Multilateral Convention. Their estimate is that more than 40% of the groups likely to be affected by Amount A of Pillar One are US-based, making the US indispensable. The prospects of the US Senate consenting to the Multilateral Convention are by no means assured.

The second sign of trouble is evident the decision to conduct yet more consultation on Amount B of Pillar One. Amount B has been relatively neglected in the prior BEPS work and this omission now seems to be causing delays, hindering the finalising of the work.

And in the same vein, the two-pronged approach to the STTR hints at disagreement. The invitation to developed countries to implement the STTR through amending bilateral treaties, rather than wait for the Multilateral Instrument, points to some doubts about the prospects for landing the Multilateral Instrument.

The Outcome Statement adds no new details or changes to the other components of Pillar Two: the domestic minimum top-up tax, the income inclusion rule or the under-taxed profits rule. These happen to be of most concern to the developed world and are apparently on track for implementation: the Statement notes, the “global minimum tax framework under Pillar Two is already a reality, with over 50 jurisdictions taking steps towards implementation.”

Toby Eggleston
Toby Eggleston
Partner, Melbourne
+61 413 151 183
Nick Heggart
Nick Heggart
Partner, Perth
+61 437 001 229
Ryan Leslie
Ryan Leslie
Partner, Melbourne
+61 418 186 499
Graeme Cooper
Graeme Cooper
Consultant, Sydney
+61 2 9322 4081
Jinny Chaimungkalanont
Jinny Chaimungkalanont
Partner, Sydney
+61 418 479 417

Tax Insight: Revised rules for non-deductible royalties

Treasury has released a revised version of draft provisions which will deny large Australian businesses deductions for payments connected with intangibles made to associates in low tax jurisdictions. The revisions were made in light of submissions on the March version of the rules, and according to the announcement, will “better achieve the policy intent.” This Tax Insight looks at some of the things that have – and haven’t – changed. Our Tax Insight on the March version of the rules is available here.

Some key changes

Interaction with royalty withholding tax and CFC rules

These provisions are directed at multiple targets – protecting royalty withholding tax where royalty-like payments do not meet the definition of “royalty,” stopping inappropriate access to the reduced rates of royalty withholding tax provided in our tax treaties, and stopping the leakage of income into tax havens. In theory, therefore, they should not be enlivened if:

  • the payment was actually a “royalty” as defined and so attracted royalty withholding tax on the way out of Australia, or
  • Australia already negates the benefit of the low tax rate enjoyed in the tax haven under our CFC rules – that is, the payment is attributable income of an Australian resident shareholder of a CFC, or
  • the rest of the world negated the low tax rate enjoyed in the tax haven through the imposition of Pillar Two measures. (Under the proposed law, one part of the definition of a “low corporate tax jurisdiction” is where a country’s corporate tax rate is below 15%, which is the rate that countries which enliven Pillar Two rules will be visiting upon income earned in the haven.)

None of these exceptions existed in the March version of the rules.

A new provision has been added in the June version to address the first point: the amount of deduction being denied will be reduced to the extent of any royalty withholding tax remitted to the ATO on the payment. The mechanics of the reduction converts the amount of royalty withholding tax paid back into the amount of a deduction. So, for example, remitting $30 of royalty withholding tax on a $100 royalty payment, will immunise the entire $100 from being made non-deductible.

But the calculation subverts the benefits of Australia’s treaties, whether or not there is any “abuse” of the treaty involved: if the payer remits only $10 of royalty withholding tax on a $100 royalty payment, only $33 is immunised and the remaining $67 is still non-deductible. (Indeed, it is hard to see why the regime needs to apply at all to payments which have survived scrutiny under the multiple tests protecting our treaties from abuse: the purpose test in the royalties article, the principal purpose test and limitation on benefits clauses.)

And it is worth noting, the provision does not take into account the impact of any royalty withholding tax levied by an intermediary country through which the payment might flow prior to reaching the “low corporate tax jurisdiction” destination.

With regard to the second point, the revised draft now addresses the scenario where Australia negates the benefit of the low rate enjoyed in the tax haven under our CFC rules. The income will not be regarded as derived in a “low corporate tax jurisdiction” to the extent that the amount is included in the assessable income of the Australian resident shareholder under our CFC rules.

Interaction with Pillar Two

But the Government has not yet been willing to address the third point: the significance of the Pillar Two measures on this rule. Under both versions of the rules, whether or not a country is a “low corporate tax jurisdiction” depends on “the rate of corporate income tax under the laws of that foreign country.” So even if the rest of the world applied Pillar Two measures to impose an effective 15% rate and negate the low tax rate enjoyed in the tax haven, that tax will not count (although a domestic minimum tax levied in the tax haven might count, depending on how it is implemented).

The announcement accompanying the revised draft rules said only, “the Government is further considering interactions of the intangibles measure with global minimum taxes and domestic minimum taxes.”

Earned in a “low corporate tax jurisdiction”

One of the key requirements for triggering the new rule is that an associate of the payer derives the relevant income (income from exploiting an intangible asset) “in a low corporate tax jurisdiction …” Our previous Tax Insight highlighted the confusion in the definition of “low corporate tax jurisdiction”: at some places, the drafting refers to the headline corporate tax rate, at other places it refers to the rate of tax imposed on a particular class of income, and at other places it refers to matters which go to the effective rate being paid by a particular taxpayer. The revised Exposure Draft makes two changes aimed at clarifying the confusion about when this test is met.

(a)   A “low corporate tax jurisdiction …”

First, the revised Draft and the Explanatory Memorandum try to be more explicit that the test of whether a country is a “low corporate tax jurisdiction” is determined by looking at the headline rate: it is the top headline rate of corporate tax, applicable to the largest taxpayers, in respect of income derived in the ordinary course of carrying on business, and disregarding any concessional rates applicable to particular industries [eg, oil and gas] or income or taxpayers, and disregarding the effect of their personal circumstances [the impact of deductions, tax credits, tax losses, and so on].

These changes should mean that a country will not be a “low corporate tax jurisdiction” just because there is no income tax on a particular class of income or if there are concessions for particular industries. But the matter is not clear because the drafting is still inconsistent: on the one hand, the legislation says to, “disregard the effect of … exemptions for particular types of income …”; on the other, it insists that only the lowest rate is considered if, “there are different rates of income tax for different types of income …”  Taken literally, this would lead to the absurd outcome, that if (say) income from oil and gas is exempt, the national headline rate applies, but if there is a low rate on oil and gas, it applies instead of the headline rate.

Two examples in the Explanatory Memorandum which go some way to clarifying the confusion, though one must always treat EMs with caution:

  • the first example says a country is a “low corporate tax jurisdiction” if the headline rate on business income is 10%, even if passive income [ie, the payment leaving Australia] might be taxed at 22%;
  • the second example says a country is not a “low corporate tax jurisdiction” if the headline rate is 20%, even if (say) manufacturing income is taxable at 10% and capital gains are not taxed.

But the problem for the drafters is obvious: deciding whether a country is, as a whole, satisfactory or not, must start with the tax rate, but it can’t really stop there if the tax base is thoroughly riddled with exempt items of income.

(b)   “Subject to foreign income tax”

The second change is a new provision added just for the purposes of the royalties measure. It says, income will not be regarded as derived in a low corporate tax jurisdiction to the extent that, “the income is, or will be, subject to foreign income tax at a rate of 15% or more …”  This provision picks up the definition of “subject to foreign income tax” from the anti-hybrid rules: the amount is “subject to foreign income tax” if it is included in the tax base of a foreign country.

It is tempting to see this provision as intended to counter the first example: even though a country as a whole is a “low corporate tax jurisdiction” because its headline rate is 10%, the impact of the new rule would be switched off if this income is “subject to foreign income tax” at a rate of 22%. Unfortunately, there is no indication in the Explanatory Memorandum about just what this core part of the new provision achieves.

Instead, the Explanatory Memorandum focuses on another part of the new provision: what happens when some elements of the definition of “subject to foreign income tax” are removed when the definition is employed in the royalties measure. In the anti-hybrid rules, an amount is not viewed as “subject to foreign tax” if some third country’s anti-hybrid rules are triggered, or if the amount is subject to a foreign State or municipal tax. These exclusions are switched off for the royalties measure and so amounts taxed under anti-hybrid rules or by State and municipal taxes will be viewed as “subject to foreign income tax.”


Another change introduced in the June version is to double the penalties that are triggered if a tax shortfall arises (because of a failure to take reasonable care, recklessness or intentional disregard of the provisions) with respect to the application of these provisions. The separate penalty for making a false and misleading statement, even where no shortfall arises, is also doubled. The doubling of these penalties is in addition to the automatic doubling of penalties that already applies for a significant global entity.

Some things haven’t changed

Some key terms used in the original draft have not been changed. In particular, the revised draft does not change some of the key imponderables:

  • when a payment is “attributable to” a right to exploit an intangible asset;
  • how to apportion a single payment if it is both “genuinely made as consideration for other things …” as well as for “a right to exploit an intangible asset …”;
  • when does the transaction involve “a right to exploit an intangible asset” that turns out to be “a right in respect of, or an interest in, a tangible asset” or land or equity interests or a financial arrangement;
  • when an acquisition happens under “an arrangement” between the taxpayer and its associate, or under a “related arrangement”;
  • when an arrangement produces the “result” that income is being derived in a low corporate tax jurisdiction.

Nor has the Government accepted submissions that these rules should explicitly state they are enlivened only where there is evidence of a purpose of avoiding tax. The Explanatory Memorandum refers to the measure as “an anti-avoidance rule designed to deter SGEs from avoiding income tax …” but the rules do not depend on any finding about purpose or intention. The only comfort comes from passages in the Explanatory Memorandum which say the rules don’t apply to transactions and structures which are “genuine” though it is hard to see how that comes about.

Further consultation and start date

While Treasury has revised its Exposure Draft, there is no indication in the announcement about a process for further consultation – the period for consultation on the March version closed in April and there is no public evidence of any extension or restart. Releasing the document as an Exposure Draft suggests Treasury may be willing to receive feedback on this text, but the absence of an explicit process suggests otherwise.

Even though the legislative process hasn’t even started – the Government has yet to introduce a Bill into Parliament – the Government has decided not to delay the commencement of the new provisions. Consequently, these rules – whatever they may end up saying – will apply to any payments made after 1 July 2023, if they are enacted as currently written.

Key contacts

Hugh Paynter
Hugh Paynter
+61 2 9225 5121
Toby Eggleston
Toby Eggleston
Partner, Melbourne
+61 413 151 183
Ryan Leslie
Ryan Leslie
Partner, Melbourne
+61 418 186 499
Nick Heggart
Nick Heggart
Partner, Perth
+61 437 001 229
Graeme Cooper
Graeme Cooper
Consultant, Sydney
+61 2 9322 4081
Jinny Chaimungkalanont
Jinny Chaimungkalanont
Partner, Sydney
+61 418 479 417

Tax Insight: Government Introduces Thin Capitalisation Changes

The Government has introduced into the House of Representatives the Treasury Laws Amendment (Making Multinationals Pay Their Fair Share – Integrity and Transparency) Bill 2023 to enact its proposed changes to Australia’s thin capitalisation regime. This Tax Insight gives a brief summary of the key points; a fuller analysis will follow.

Thin capitalisation changes

So far as thin capitalisation is concerned, the Bill will enact the key components of the policy which the ALP took to last year’s election and released for consultation earlier this year. See our Tax Insight at

For general investors:

  • The current safe harbour for general investors (the level of debt must not exceed 60% of the value of assets) will be replaced by test which limits debt deductions to no more than 30% of tax EBITDA.
  • Where interest is denied under this test, the deficiency may be carried forward for up to 15 years.
  • The current option which allows gearing of the Australian entity up to the gearing level of the worldwide group will be replaced by a test based on the adjusted earnings ratio of the worldwide group.
  • The arm’s length debt test has been replaced by an external third party debt test.
  • The new tests start from 1 July 2023.

There have been modifications to some of the detail in the Exposure Draft, but the key design elements remain.

Section 25-90 deferred, but …

An Exposure Draft of the Bill released for comment earlier this year had included a proposal to repeal s. 25-90. See our Tax Insight at

This measure has been omitted from the Bill, but it is not dead. The Explanatory Memorandum to the Bill says, “stakeholder concerns regarding section 25-90 were considered by Government, with the proposed amendment deferred, reflected in its removal from the final legislation, to be considered via a separate process to this interest limitation measure.”

And the deferral of s. 25-90 came at a price. The Bill includes new “debt creation” rules (based on the former Div 16G ITAA 1936) which were not released for consultation. The Explanatory Memorandum to the Bill says, the “targeted debt creation rules were progressed in its place.” The measure will disallow debt deductions, “to the extent that they are incurred in relation to debt creation schemes”. The Bill outlines 2 cases:

  • an entity borrows to acquire an asset from an associate, and
  • an entity borrows from an associate to fund a payment it will make to that entity or another associate.

These rules will operate independently of the thin capitalisation regime, as a discrete measure to deny interest deductions. They will require careful and detailed analysis.

Other measures in the Bill

The Bill also contains amendments to the Corporations Act 2001 to require Australian public companies (listed and unlisted) to disclose itemised information about subsidiaries in their annual financial reports. This requirement will apply to financial reports for financial years starting on and after 1 July 2023.

Status update

The Government also took the opportunity to give an update on the work toward meeting some of its election commitments, re-announced in the October 2022 Budget, and previously released for consultation. Consequently, a large part of the Explanatory Memorandum to the Bill has nothing to do with measures in the Bill; it is devoted to explaining what is happening elsewhere. These measures include:

Measure Update
MNEs will be required to publish tax information drawn from country-by-country (CbC) reports which will be made public

[See our Tax Insight at]

“… four additional data disclosures – related party expenses, the effective tax rate disclosure and the two intangible assets disclosures – [have been] removed from the proposed option

The proposed [measure is] deferred … by 12 months, to apply from 1 July 2024 …

While the disaggregated CbC reporting is intended to support meaningful improvements to tax transparency disclosures, there is a recognition that it does depart from the EU and OECD approaches, and that further consultation with industry may be beneficial on this element of the measure (and the measure more broadly).”

Tenderers for Australian government contracts worth more than $200,000 must disclose their country of tax domicile. “This element does not require legislative amendments and will instead be implemented via administrative changes to the Commonwealth Procurement material”
Denying a deduction for payments made by a SGE to a related party for the exploitation of an intangible asset, where the arrangement leads to income derived in a low or no-tax jurisdiction.

[See our Tax Insight at]

“The option is the preferred option … The option applies to payments made from 1 July 2023.”


Toby Eggleston
Toby Eggleston
Partner, Melbourne
+61 413 151 183
Ryan Leslie
Ryan Leslie
Partner, Melbourne
+61 418 186 499
Nick Heggart
Nick Heggart
Partner, Perth
+61 437 001 229
Graeme Cooper
Graeme Cooper
Consultant, Sydney
+61 2 9322 4081
Jinny Chaimungkalanont
Jinny Chaimungkalanont
Partner, Sydney
+61 418 479 417

Tax Insight – Australia’s Second Income Tax

The recent Budget contained only a few surprises but one important announcement was the confirmation of what many had already suspected: Australia will soon have a second income tax. In fact, it will start in just over 6 months.

Australia will enact a so-called “qualified domestic minimum top-up tax” (“QDMTT”) as part of the suite of measures to enact the OECD’s Pillar Two proposals into Australian law, along with the proposed Income Inclusion Rule (“IIR”), Undertaxed Profits Rule (“UTPR”) and possibly a Subject to Tax Rule. Those rules will be enacted to set out Australia’s claim to any low-taxed profits earned by foreign companies in case other countries forget to enact their own QDMTT. But they are almost certainly going to become irrelevant verbiage and very quickly: the main game has already shifted to countries enacting their own QDMTT, allowing them to get in first and switch off other countries’ tax claims. The impetus to do this has only increased with the threat from the Republican members of the US Congress Ways and Means Committee to retaliate against the companies of any country which tries to collect tax on the profits of US companies under the UTPR.

Consequently, speculation has now shifted to just what Australia’s second income tax will look like. The optimists are hoping Treasury will closely follow the design of the IIR and UTPR: the QDMTT will be designed in such a way that it leaves no room for other countries’ IIRs and UTPRs to operate, but nothing more. The pessimists suspect Treasury won’t want to stop there: Treasury will take the opportunity of imposing a brand new tax to design one unpolluted by any of Treasury’s bêtes noires. The optimists reply: a tax that departs from the design of the IIR and UTPR won’t meet the test of a QDMTT so Treasury will be constrained from being too ambitious. The pessimists respond: the new tax will still switch off foreign governments’ claims as long as it is a tax on profits (“a covered tax”); it doesn’t also need to meet the definition of a QDMTT to be effective in blocking foreign governments.

This issue is currently playing out in the debate around the US’ second alternate corporate minimum tax which began operation on 1 January: people are asking whether it is a QDMTT or not. But the answer to that question is not especially important if it is instead either a second covered tax or if it increases the amount of an existing covered tax.

The same question arises for Australia: will Treasury aim for (i) a tax that is a very close approximation of the IIR and UTPR rules, or (ii) a tax that takes the IIR and UTPR as the model but has some important departures, or (iii) a tax that abandons any attempt to be a QDMTT and aims instead to be a second covered tax?

Option (i) would be the least ambitious approach – to have the QDMTT simply mirror the IIR and UTPR design.

But there are reasons to suspect the tax Australia will end up with is option (ii) – a tax which meets the QDMTT definition, but ends up departing from the model and is more ambitious.

First, the OECD is already encouraging countries to be more aggressive when it comes to their QDMTT. For example,

  • the IIR and UTPR are only triggered for very large taxpayers: MNE groups with revenue exceeding €750m. But when it comes to a QDMTT the OECD invites countries to waive this requirement, “… the application of a QDMTT could be extended to groups … that are not within the scope of the [IIR and UTPR] Rules because their revenues are below the EUR 750 million threshold”;
  • in the same vein, the IIR and UTPR are only triggered for groups with cross-border operations: either a subsidiary or PE located in a jurisdiction different from the ultimate parent. The OECD invites countries to waive this requirement as well: “… a QDMTT could also apply to purely domestic groups, i.e. groups with no foreign subsidiaries or branches”. Some countries will take up this invitation. For example, the EU has decided it will apply the Pillar Two measures to groups that only operate domestically and have no foreign subsidiaries or branches;
  • the IIR does not apply if the low-taxed entity in the group is the ultimate parent. The OECD says a QDMTT should apply to the ultimate parent: “… a QDMTT should impose a Top-up Tax on … all domestic Constituent Entities, including the domestic Parent Entity”;
  • the IIR and UTPR have an active income exemption (the “substance-based income exception”) which renders some income immune from top-up tax. The QDMTT does not need to have such an exemption: “a QDMTT is not required to have a substance carve-out”;
  • the IIR and UTPR have a de minimis exemption but a QDMTT is not required to have a de minimis exclusion: “a QDMTT is not required to have a De minimis exclusion … in order to be considered functionally equivalent to the GloBE rules”;
  • the base for the IIR and UTPR is set by the accounting standards used for preparing the consolidated accounts of the ultimate parent entity. The tax base of a QDMTT can be based on local financial accounting standards: “… a jurisdiction may require income or loss for the jurisdiction to be computed using an Authorised Financial Accounting Standard that differs from the one used in the Consolidated Financial Statements”;
  • the base of the IIR and UTPR reverses deductions for fines and penalties above €50,000 if they are deducted in the financial accounts. A QDMTT can reverse a deduction for even smaller fines or penalties: “… a jurisdiction that does not permit deduction of fines and penalties in any amount under its corporate income tax (CIT) can apply the same standard under its QDMTT”;
  • the amount of top-up tax which will have to be paid under the IIR and UTPR requires a decision about which foreign taxes will count toward the minimum. But the OECD is quite happy if a QDMTT does not treat as a covered tax a tax which would be counted toward reducing the IIR or UTPR: “the determination of Adjusted Covered Taxes needs to be the same or more restrictive”;
  • the IIR and UTPR are subject to transitional and permanent safe harbours. The OECD says safe harbours should also exist in a QDMTT but they do not need to be exactly the same: “the Inclusive Framework will undertake further work on the development of a QDMTT Safe Harbour …”
  • the IIR and UTP rules contain many elections which groups may choose in order to mitigate some requirement or other. When it comes to a QDMTT, the OECD says the QDMTT should generally provide a corresponding election, except, “a QDMTT that does not provide for certain elections, for example GloBE Loss Election, may be functionally equivalent.”

We could go on, but these examples are sufficient to show the OECD is trying to sit on 2 horses at the same time. On the one hand, it tells countries, “a minimum tax must follow the architecture of the [IIR and UTPR] rules using mechanisms that are substantially the same as those used to calculate the effective tax rate and top-up tax payable …”

On the other hand, it tells countries that being more ambitious won’t disqualify their tax as a QDMTT: “Some degree of customisation of a QDMTT in each jurisdiction is to be expected [and] variations in outcomes between the minimum tax and [IIR and UTPR] Rules will not prevent that tax from being treated as a QDMTT if those variations systemically produce a greater incremental tax liability …” The message is clear: a QDMTT must not collect less tax than the IIR or UTPR; collecting more is no problem.

A second reason to expect divergence is, there are aspects of IIR and UTPR design which serve no purpose in a QDMTT. For example,

  • the IIR and UTPR rules devote a large amount of legislative effort to rules that determine which country will get the top-up tax if the IIR is enlivened [which entity is the “ultimate parent entity”, an “intermediate parent entity,” or “partly-owned parent entity”], or if the UTPR is enlivened [the allocation key rules]. None of that detail needs to exist in a QDMTT since all of the tax to be paid under the QDMTT is meant to be paid to the Australian government and no-one else;
  • the IIR and UTPR rules devote a large amount of legislative effort to determining who is an MNE group and whose income and losses can be aggregated [netting is prohibited with entities that are investment entities, joint ventures and minority-owned constituent entities]. Treasury won’t be able to use the tax consolidated group definition from the Australian consolidation regime since it is under-inclusive compared to the IIR and UTPR so there will need to be both more entities included and more entities excluded. The simplest solution may simply be to treat all affected entities as stand-alone entities and not allow netting across any entities in the QDMTT.

It is inevitable the QDMTT will not simply mirror the IIR and UTPR design: Australia can be more ambitious in designing a QDMTT to tax Australian companies than it can be when applying the IIR or UTPR to claim tax from foreign companies.

Option (iii) would be the most ambitious approach: Treasury might abandon any effort at meeting the definition of a QDMTT (or simply not care whether its variations have strayed too far from the definition) and rely instead on the new tax meeting the definition of “covered tax”: a tax “recorded in the financial accounts of a Constituent Entity with respect to its income or profits …” The Commentary to the Model Rules already contemplates, and explains how to handle, “a domestic minimum tax that is not a Qualified Domestic Minimum Top-up Tax but that meets the definition of a Covered Tax …” The opportunity to write the corporate tax Treasury always dreamt of might prove irresistible. There is no down-side in enacting radical departures from the IIR and UTPR provided the tax is still a tax on profits.

It seems inevitable that Australia’s second income tax will end up departing from the parameters set in the IIR and UTPR regimes, but we won’t know just how far until the legislation is released, hopefully later this year.

Toby Eggleston
Toby Eggleston
Partner, Melbourne
+61 413 151 183
Ryan Leslie
Ryan Leslie
Partner, Melbourne
+61 418 186 499
Nick Heggart
Nick Heggart
Partner, Perth
+61 437 001 229
Graeme Cooper
Graeme Cooper
Consultant, Sydney
+61 2 9322 4081
Jinny Chaimungkalanont
Jinny Chaimungkalanont
Partner, Sydney
+61 418 479 417

Tax Insight – Another public tax reporting obligation

Treasury has released for consultation an Exposure Draft of new rules which will require multinational groups with an Australian presence to publish information about their tax and financial position for all the jurisdictions in which the group operates. Submissions on the Exposure Draft are due by 28 April 2023.

1. Background

The Exposure Draft just released is the latest development in the campaign being waged in Australia and around the world to deploy publicity as a tool to discourage tax avoidance by MNEs.

Australia already has legislative regimes which encourage or require entities to report information about their tax affairs to the public:

  • in February 2013, the Assistant Treasurer announced that the Government would legislate “to improve the transparency of Australia’s business tax system,” an announcement which led eventually to the annual ATO Report of Entity Tax Information; and
  • in May 2015, the Treasurer commissioned the Board of Taxation to develop a Voluntary Tax Transparency Code to encourage large multinationals to disclose tax information.

While these measures were being legislated, many large taxpayers decided to act unilaterally and began publishing a report on taxes paid around the world as part of their suite of annual reports.

Meanwhile, the OECD was working on item 13 of the BEPS Action Plan. The Final Report on Action 13, released in October 2015, required large groups to prepare a country-by-country report with aggregate data on income, profit, taxes paid and economic activity in each of the jurisdictions in which the group operates. This data would be shared with the tax administrations of various jurisdictions on a confidential basis. These requirements were legislated in Australia with effect from 2016.

But apparently, these measures were seen as insufficient:

  • the ATO Tax Information Reports and reports under the Voluntary Tax Transparency Code are publicly accessible but they focus on Australian tax;
  • a taxpayer’s own tax and financial reports are not mandatory, and reported data is selected by the group, making comparisons between groups difficult;
  • country-by-country reports are more granular and uniform, but they are only shared between revenue authorities and (by design) are not publicly available.

Consequently, in the lead up to the 2019 election the ALP promised, “to introduce public reporting of country-by-country reports, ensuring the release of high-level tax information about where and how much tax was paid by large corporations …” This proposal was repeated for the 2022 election with a promise to introduce, “transparency measures including reporting requirements on tax information, beneficial ownership [and] tax haven exposure …”

In August 2022, Treasury released a Consultation Paper, Government election commitments: Multinational tax integrity and enhanced tax transparency seeking submissions on several transparency measures including:

  • mandatory reporting of material tax risk to shareholders of Australian public companies (an Exposure Draft of amendments to the Corporations Act to give effect to this measure was released in mid-March); and
  • public reporting of information drawn from country-by-country reports (which is the focus of this Exposure Draft).

This Exposure Draft is the next step in this process. It is meant to lead to standardised tax and financial information, being collected by the ATO and then made freely available to the public worldwide, detailing a group’s worldwide operations disaggregated for each jurisdiction.

2. The new requirements

Who is affected

The obligation to report information to the ATO is imposed on the parent entity of a country-by-country reporting group required to prepare a country-by-country report, provided the parent is an Australian resident, some member of the group is a resident, or a member of the group has a PE in Australia. The references to the country-by-country reporting rules will link the measure to definitions already set out in the transfer pricing rules and will limit the measure to large multinational groups.

Foreign parent companies will have to report to the ATO if they have subsidiaries or branches in Australia.

The obligation extends to parent entities which are trusts if they have a corporate trustee and to partnership if all the members are corporations.

An exception is available for government-related entities, and the ATO has the power to exclude individual entities and classes of entities.

What to report

An affected entity is obliged to report the itemised information, “in respect of each jurisdiction in which the country-by-country reporting group operates.” The data is thus aggregated for all entities operating in the same jurisdiction to mirror the reporting requirements of the country-by-country report.

While the Exposure Draft alludes to the existing country-by-country reporting regimes, the information which must be reported is more extensive than country-by-country reports require and so cannot be drawn directly from a group’s report. Rather, the legislation lists particular pieces of information which must be reported, and says the information, “must be based on amounts as shown in the audited consolidated financial statements [of the parent entity] …”

The required information includes:

  • a description of group’s main business activities in each jurisdiction;
  • the number of people employed in the jurisdiction as at the end of the income year;
  • the book value of tangible and intangible assets at the end of the income year, other than cash;
  • the gross revenue from transactions with unrelated parties, as well as the revenue and expenses from transactions with related parties (if they are not resident in the same jurisdiction);
  • the profit or loss before income tax, the income tax actually paid and any income tax expense accrued for the current year;
  • a calculation of the effective tax rate and an explanation of the difference (if any) between the tax expense accrued and the tax that would payable if profit before tax was liable to “income tax [at the] rate applicable in the jurisdiction”; and
  • the currency used in the report.

In addition to the granular information, the Exposure Draft requires the parent to provide, “a description of the … group’s approach to tax.” This concept already appears in the voluntary Tax Transparency Code, and reports usually say things like, “we are committed to complying with all relevant tax legislation; we will not adopt tax positions that are considered aggressive; we will work co-operatively and openly with relevant tax authorities …” and so on. Reports under the new regime will likely say much the same thing.

An entity must notify the ATO of corrections if it becomes aware of errors in the original report.

How to report

The Exposure Draft goes through a curious charade. The obligation is imposed on an affected entity to “publish” the information required by the amendment, but an entity can publish the information only by giving, “a document containing the [required] information to the Commissioner in the approved form.” The Commissioner then makes “the information in the document available on an Australian government website.”

When to report

An entity affected by the measures has up to 12 months after the end of the income year to give the required information to the ATO, although there is a power for the ATO to vary this time limit, presumably for groups with later country-by-country reporting deadlines.

Corrections must be notified to the ATO “as soon as practicable” after the entity becomes aware that its original document contains an error.


The Exposure Draft proposes amending the Taxation Administration Act 1953 to add a new strict liability offence where an affected entity fails to comply with its obligation to give the information to the ATO in the required manner. Just how this penalty would be enforced against a non-resident parent is not obvious, but the issue is unlikely to arise.

3. Comment

The result of the process looks very much like the ATO is collecting information about the worldwide operations of MNEs from country-by-country reports, and then publishing it to the world. The Final Report on BEPS Action 13 insists:

Tax administrations should take all reasonable steps to ensure that there is no public disclosure of confidential information … and other commercially sensitive information contained in the [country-by-country documents and] should also assure taxpayers that the information presented in transfer pricing documentation will remain confidential.

The Exposure Draft acknowledges that the information in the country-by-country report is subject to strict confidentiality and cannot be publicly disclosed.

So the Government attempts to sidestep this obligation by simply requiring companies to provide the information in a “separate public reporting obligation”. No doubt this two-step charade [the entity publishes its own information; the ATO just helps out], and the statement that the information uses amounts shown in the audited consolidated financial statements, is meant to convince other governments that the ATO is not simply publishing the contents of country-by-country reports, in breach of the secrecy obligations. One suspects, other governments may see things differently.

4. Dates

Submissions on the Exposure Draft are due by 28 April 2023.

If enacted in its current form, the measure would require taxpayers to report to the ATO information about income years starting from 1 July 2023.

Toby Eggleston
Toby Eggleston
+61 3 9288 1454

Nick Heggart
Nick Heggart
+61 8 9211 7593

Ryan Leslie
Ryan Leslie
+61 3 9288 1441

Hugh Paynter
Hugh Paynter
+61 2 9225 5121

Jinny Chaimungkalanont
Jinny Chaimungkalanont
+61 2 9322 4403

Graeme Cooper
Graeme Cooper
+61 2 9322 4081


Tax Insight – New Limits on Deducting Payments Involving Intangibles

Treasury has released for consultation an Exposure Draft of new provisions which will, in some cases, deny large Australian businesses a deduction for amounts connected with intangibles if paid to associates. If enacted in their current form, the new rules would apply from 1 July 2023. Submissions on the Exposure Draft are due by 28 April 2023.

1. Background

One of the tax policies which the ALP announced in its unsuccessful campaign for the 2019 election was a proposal to, “stop multinationals from getting a tax deduction when they unfairly funnel royalty payments to arms of their own company that pose a multinational tax risk.” The announcement listed the sources of “tax risk” as treaty shopping, funnelling income into tax havens, holding patents in countries with patent box regimes and negotiating “a sweetheart deal” with another country.

In April 2022, in the campaign for the 2022 election, the ALP announced a similar proposal to limit, “the ability for multinationals to abuse Australia’s tax treaties when holding intellectual property in tax havens.”

In August 2022, Treasury released a Discussion Paper outlining the design of, “a new rule limiting MNEs’ ability to claim tax deductions for payments relating to intangibles and royalties that lead to insufficient tax paid.” (This Paper is examined in more detail in our Tax Insight available here.)

Treasury has now released an Exposure Draft of provisions to give effect to the policy. The detail has evolved significantly in the last 4 years. The Exposure Draft is available here.

2. Key Elements

The new rule will deny a deduction to an Australian entity for a payment (or crediting an amount or incurring a liability to make a payment) where 5 key conditions are met:

  1. the payer is a significant global entity;
  2. it is making the payment to an associate, whether directly or indirectly;
  3. the payment is, “attributable to a right to exploit an intangible asset”;
  4. as a result of some arrangement, income from the exploitation of an intangible is derived by the recipient or another associate; and
  5. that income is derived in a low corporate tax jurisdiction.

3. Some of the pitfalls

At first glance, the intended scope of the section seems easy grasp, but the apparent simplicity of the drafting hides many traps. Most of the complexity arises from:

  • asserting that a transaction can be re-characterised to reveal a payment for a right to use an intangible; and
  • asserting that the Australian taxpayer is to be denied a deduction because another transaction, to which the taxpayer is not a party, occurs between other entities in the same corporate group, which shifts income into a low tax jurisdiction.

Royalties versus deductible “payments”

While the focus of the 2019 announcement was on limiting the deductibility of “royalty payments” made by MNEs, by 2022, Treasury’s Discussion Paper revealed the policy had expanded to include payments which were not technically “royalties” and did not trigger royalty withholding tax. The Paper gave as examples, payments for goods, payments for services or payment of management fees.

The Exposure Draft adopts this broader scope. The Explanatory Memorandum says the drafting is meant to extend to, “a payment … made for other things, such as services or tangible goods …” Hence, an Australian SGE may find deductions are being denied for the cost of trading stock or management fees or payments for services.

However, it seems the drafting is not enlivened for payments which are not deductible, but rather form part of the cost of depreciating assets or the cost base of CGT assets.

Embedded royalties: payments “attributable to” a right to exploit an intangible asset

In order for a deductible payment to be in jeopardy, it is necessary that “the [deductible] payment is attributable to a right to exploit an intangible asset …”

Obvious examples would include payments for the right to use a copyright, patent or registered design in the Australian entity’s operations, payments for the right to display trademarks and logos in marketing activities, payments for rights to broadcast radio and TV content, and so on. The drafting extends the definition to include payment for rights to exploit things which aren’t property at law, such as gaining access to confidential information. The phrase “attributable to” is probably meant to capture payments for supporting services in connection with the licensing of IP or revealing information. In all these cases, the Australian entity acquires rights in connection with an intangible which is owned (and remains owned) by another. The Explanatory Memorandum says the word “exploit” is meant to be broader than “use,” but the essence of the test still requires that the Australian entity gets a limited right to exploit something owned by another.

Which is why the attempt to capture embedded royalties is so fraught. In what circumstances can a payment for trading stock or management services be re-labelled as “attributable to” a right to exploit an intangible? The Explanatory Memorandum offers no explicit guidance about how and when a payment is “attributable to” a right to exploit an intangible. The example in the Explanatory Memorandum involves a taxpayer who distributes products under a distribution agreement. It pays for management and other services it acquires under the agreement; it is allowed (at no cost) to display logos etc when marketing the goods. The Example asserts, without explanation, that a portion of the management fee paid to the other party is attributable to exploiting the IP. The Example is equally reticent about how to quantify the non-deductible portion of the payment saying simply that, “to the extent that the payment of these fees is attributable to the right to exploit the trademark …” it will be non-deductible.

The new rule does not apply to payments connected to every kind of intangible. Rather it is focused on payments for an intangible right over another intangible asset. Hence, the new rule does not apply if the payment is made for a right to use or exploit:

  • a tangible asset (eg, rentals under an equipment lease); or
  • land (eg, payments of rent, or fees for a licence over land).

And the new rule does not apply to rights over every kind of intangible asset: a right to exploit a financial arrangement affected by the TOFA rules is excluded (eg, interest, dividends, gains or losses in relation to some debt or equity interests).

And, one would have thought, the rule should not extend to payments to buy rather than exploit – whether to buy goods, or land or an intangible – though the drafting says it will be a payment to “exploit” an intangible if the payment is being made to allow the entity to sell or license the intangible. So, it is hard to see how a payment to buy branded goods could involve a disguised payment to exploit the brand. TA 2018/2 Mischaracterisation of activities or payments in connection with intangible assets, which foreshadowed these rules, says it does not apply to, “resellers of finished tangible goods where the activity of reselling the goods involves an incidental use of a brand name that appears on the goods and related packaging.” One would hope the final legislation will include a similar limitation in these rules.

But this issue of embedded royalties is clearly a major concern to the ATO which is trying to challenge them under current law. It is understood the ATO is arguing, in its case against Pepsi, that the transactions involved there were amenable to challenge under the diverted profits tax. And the issue was the major focus of TA 2018/2 Mischaracterisation of activities or payments in connection with intangible assets. The TA says the transactions outlined in the TA may contain a “royalty” liable to withholding tax, may trigger transfer pricing rules or might trigger Part IVA. The new provision will add another option to the list.

Payments pursuant to “arrangements”, involving multiple entities

In order for a deductible payment to be in jeopardy, it is necessary that:

  • “the [deductible] payment is attributable to a right to exploit an intangible asset …”; and
  • the taxpayer or an associate acquires the intangible asset, or a right to exploit the intangible asset or actually exploits the intangible asset; and
  • that acquisition happens under
    • the arrangement between the taxpayer and its associate which conferred the right to exploit the intangible asset, or
    • a related arrangement (the parties to which are not itemised).

So the drafting contemplates at least one, and possibly several arrangements:

  • the plain vanilla case is: the taxpayer pays a royalty to an associate in a tax haven for a right to exploit an intangible asset and it gets the right it bargained for;
  • the drafting easily captures back-to-back arrangements: the taxpayer pays a royalty to an associate not in a tax haven for a right to exploit an intangible asset, and associate pays another fee to an associate which is in a tax haven for the rights which it has licensed to the Australian entity. The legislation expressly refers to the Australian entity making the payment “to the recipient directly or through one or more other entities”;
  • a more complex case is: the taxpayer pays a royalty to an associate not in a tax haven for a right to exploit an intangible asset; another associate of the taxpayer which is in a tax haven owns or licenses the same or related intangible under an agreement and derives income from another associated entity. The only requirement for this scenario is that (i) the arrangement between the Australian payer and the recipient and (ii) the arrangement between the other entities are somehow “related.”

Low corporate tax jurisdiction

The final requirement for triggering the rule is that, as a result of the agreement, the recipient or some other associate derives income in a low corporate tax jurisdiction.

A “low corporate tax jurisdiction” is defined to mean:

  • a foreign country where the rate of corporate income tax is less than 15% or is nil; or
  • a foreign country which the Minister has determined offers a preferential patent box regime which does not require sufficient economic substance (a test taken from the OECD’s Final Report on BEPS Action 5).

The drafting does not appear to capture a country which simply does not have a corporate tax.

The 2019 announcement said the new regime would use the “sufficient foreign tax test” from the Diverted Profits Tax, but the test finally settled upon in the Exposure Draft is far more manageable:

  • the test focuses on the headline rate for the corporate tax of the country;
  • if the country applies progressive rates to companies, the relevant rate is the highest headline rate;
  • if the country applies different rates to residents and non-residents, ignore a rate which applies only to non-residents.

The drafting refers to “the rate of corporate income tax under the laws of [a] foreign country …” which means the test looks at the rate for the country, not the rate of tax being paid by a particular taxpayer (eg, does the taxpayer have losses or tax credits), nor the rate of tax imposed on a particular class of income (eg, is this kind of income exempt). However, the drafting does become more granular:

  • the impact of deductions, losses and tax credits are excluded;
  • if the income is exempt from tax, we are told the rate of tax on that amount is nil.

In theory, both of these matters should be irrelevant to identifying “the rate of corporate income tax under the laws of [a] foreign country …” but their presence confuses the matter, and indicates the headline rate may not be the end of the story.

Purpose of avoiding tax?

The 2019 announcement was clear that this rule would only be triggered if the taxpayer had a purpose of avoiding Australian tax: “a multinational can still get the tax deduction if the firm can substantiate to the Commissioner of Taxation that the royalty payments are not for the dominant purpose of tax avoidance.”

This test has now been abandoned in favour of a “results” test: it will be sufficient if, “the entering into of the arrangement … results in the recipient or another associate of yours deriving income” in a low corporate tax jurisdiction. So, the focus is now simply on what happened, not why it happened.

The Explanatory Memorandum says, “it is not intended for this anti-avoidance rule to inappropriately apply … where there is no tax avoidance behaviour” but the drafting contains no test of “purpose” or “intention” which might limit the inappropriate operation of the section.

This will be left to the ATO to administer, but how this issue is to be dealt with in practice is left unresolved.

4. Royalty Withholding Tax?

It is worth noting that nothing in the Exposure Draft deals with the interaction between this measure and the tax on payments which are royalties and subject to royalty withholding tax. It is conceivable that both measures will operate. In the case of a royalty paid to a non-treaty country, this might lead to an effective rate of 60%: 30% imposed on the recipient (and collected from the payer) and 30% imposed on the payer through the denied deduction.

The media releases accompanying the Exposure Draft have referred to “levelling the playing field for Australian businesses”. However, by both imposing royalty withholding tax and denying a deduction for the payer it is clear that these measures will tilt the field in favour of Australian enterprises. Whether this level of support is acceptable to our trading partners will be a watching brief. The US Treasury has already indicated its displeasure with the ATO’s expanded view of royalties as expressed in TR 2021/D4.

One would have expected the measure would not apply where the Australian payer has complied with Australian royalty withholding tax rules (including having regard to applicable tax treaties) but that suggestion has not been taken up in the current drafting.

5. Interaction with Pillar 2

Finally, there is no mention of the interaction of this measure with the OECD’s Pillar 2 project of a global minimum tax of 15%. If Pillar 2 is widely adopted (and the income inclusion rule is proposed to start for income years after 1 January 2024), then whatever justification for this measure will fall away given the income for exploiting intangibles will be picked up and taxed at a minimum of 15%.

6. Dates

The new rule will apply to payments made (or liabilities incurred) from 1 July 2023.

Submissions on the Exposure Draft are due by 28 April 2023.


Toby Eggleston
Toby Eggleston
+61 3 9288 1454

Ryan Leslie
Ryan Leslie
+61 3 9288 1441

Jinny Chaimungkalanont
Jinny Chaimungkalanont
+61 2 9322 4403

Hugh Paynter
Hugh Paynter
+61 2 9225 5121

Graeme Cooper
Graeme Cooper
+61 2 9322 4081

Tax Insight: Changes to Thin Capitalisation Rules

Treasury has released an Exposure Draft of provisions to enact the new thin capitalisation limits, to give effect to the policy which the ALP took to the last election. The Draft contains a few surprises, not least, the novel way the Exposure Draft says it meets the ALP’s promise that the new test would be based on profits “…while maintaining the arm’s length test”. This Tax Insight focuses principally on the changes affecting commercial and industrial MNEs.

1.   Key Points

  • The Exposure Draft creates a new class of taxpayers for the purposes of the thin capitalisation rules: the ‘general class investor’ which combines both foreign owned Australian entities and Australian entities with foreign operations.
  • General class investors will be subject to a new thin capitalisation test based on their adjusted earnings (30% of ‘tax EBITDA’) rather than the average value of their assets and debt. This is referred to as the ‘fixed ratio test’ and does not operate as a safe harbour.
  • Deductions denied under this test (but not the other tests mentioned below) can be carried forward for up to 15 years. Denied deductions will be lost if there is a change of ownership of a company with carry-forward denied deductions.
  • The Exposure Draft permits some taxpayers to adopt a test based on the adjusted earnings ratio of the worldwide group (in lieu of the existing worldwide gearing test).
  • The Exposure Draft repeals the arm’s length debt test for general class investors and financial entities and enacts instead an external third party debt test.
  • The treatment of the category of “financial” investors is left largely unaffected except for replacing the arm’s length debt test with the external third party debt test.
  • The treatment of Authorised Deposit-taking Institutions is left largely untouched.
  • The Exposure Draft also makes significant changes to the deductibility of interest on money borrowed to acquire shares in offshore subsidiaries (see our Tax Insight available here).
  • The Exposure Draft also adjusts the transfer pricing rules to require general class investors to demonstrate that the amount of debt they owe is not excessive (ie, even if the fixed ratio test is applied and debt deductions are less than 30% of tax EBITDA).
  • The new rules will apply for income years commencing on or after 1 July 2023 with no grandfathering of existing debt or transition period to allow for the reorganisation of existing structures.
  • Treasury has set a deadline of 13 April 2023 for submissions on the Exposure Draft.

2.   Who is affected

The principal changes affect the classes of taxpayers who used to be known as –

  • an inward investor (general) – a foreign entity with a PE or other investments in Australia,
  • an inward investment vehicle (general) – an Australian company, trust or partnership that is controlled by foreign residents, and
  • an outward investor (general) – an Australian entity that –
    • controls a foreign company, trust or corporate limited partnership,
    • carries on business abroad through a foreign PE, or
    • is an ‘associate entity’ of an outward investor.

These three groups are now combined into a single class of taxpayer – general class investors – which is made subject to the new thin capitalisation tests.

The other classes of taxpayer established in the existing rules – inward and outward financial entities and inward and outward ADIs – are also affected by the changes:

  • the Draft tightens the definition of ‘financial entity’ to limit access to the rules for financial entities on the basis that the rules for financial entities, “are generally more favourable to taxpayers than the new tests which only apply to general class investors”; and
  • the tests available to financial entities are also modified.

But provisions in the existing law which exclude certain securitisation vehicles and taxpayers with debt deductions below $2m remain unaffected (but also not indexed) by the Draft. The existing exemption for outward investing entities where Australian assets represent at least 90% of the value of total assets (on an associate-inclusive basis) is also unaffected.

Determining status. The rules now create 5 classes of entity and each class must apply the rules from the suite of rules relevant to that class. Prima facie, every entity is classified as a general entity and can be classified as financial investor or ADI only if it has that status for the whole of the income year.  (Under the prior law, status could change during a year and the relevant treatment applied to an entity for each period “that is all or a part of an income year …”)

The former distinction between inbound and outbound investors has been eliminated for general class investors but this distinction still survives for financial investors and ADIs. This creates a difficulty when an entity is both (eg, a foreign-owned Australian company with offshore operations).

3.   What is the impact of the changes

The main impact of the measures in the Draft is:

The existing test … Will be replaced by …
For general investors
Safe harbour debt amount
Arm’s length debt amount
Worldwide gearing debt amount
Fixed ratio test
External third-party debt test
Group ratio test
For inward and outward financial investors
Safe harbour debt amount
Arm’s length debt amount
Worldwide gearing debt amount
No change
External third-party debt test
No change
For inward and outward ADIs
Safe harbour capital amount
Arm’s length capital amount
Worldwide capital amount (outward investor ADI only)
No change
No change
No change

This change means Australia will employ multiple concepts for determining whether an Australian entity is thinly capitalised:

  • general investors can apply a test which looks at cash flows: interest as a percentage of earnings,
  • financial investors can continue to apply a test which looks at capital structure: the value of debt as a percentage of the value of assets, and
  • ADIs can continue to apply a test which looks at the entity’s regulatory capital: the value of its risk-weighted assets.
3.1   How the options will work for each class of taxpayer

Making an election: As the table above shows, each class of taxpayer has, and will continue to have, access to multiple tests to determine whether some or all of its debt deductions are disallowed.

Current law does not involve an explicit election: the taxpayer automatically gets the best outcome under any of the three tests. This means, the taxpayer can be wrong and not suffer – if needed, it can defend an audit relying on a method it did not actually apply. Further, the taxpayer does not need to make a formal election that it is choosing to apply one method rather than another.

The new law will work rather differently for general investors:

  • the taxpayer must apply the fixed ratio method for the income year unless it is eligible and elects to use either the group ratio test or the external third-party debt test for the year,
  • if it makes an election to use the group ratio test or the external third-party debt test,
    • the debt deduction is dictated by that election even if another method would produce a more favourable result,
    • the entity must notify the ATO in the approved form that it is making this election,
    • that election applies for the relevant year and is irrevocable for that year,
    • while a different election (or no election) can be made for subsequent years, changing methods across years can produce an unfavourable outcome if the taxpayer wants to carry forward denied deductions under the fixed ratio test to use in later years.

The existing automatic system will continue to apply to financial investors and ADIs.

One-in-all-in effect. As will be seen, for general class investors, the fixed ratio limit applies unless the taxpayer qualifies and decides to elect into either the group ratio limit or the external third party debt limit. The election is made by the taxpayer and its effect applies to all of the taxpayer’s debt. To put this the other way, a taxpayer cannot, for example, apply the external third party debt test to its external debt and apply the fixed ratio test to the remainder.

3.2   Fixed ratio test limit

The new fixed ratio test limit (based on interest as a percentage of adjusted cash flows) will replace the existing safe harbour test (based on the level of debt as a percentage of asset values) for general class investors.

The fixed ratio test disallows “net debt deductions” in excess of 30% of “tax EBITDA.”

Net debt deductions. The term “debt deductions” is used in the current law. It extends beyond interest to include some fees, rent, brokerage, stamp duty, net losses on some security arrangements and other amounts which are not technically “interest.”

The Exposure Draft also expands the existing definition of “debt deduction” to break the current link to the “debt test” in the debt-equity provisions – the existing definition requires that the expense is “incurred in relation to a debt interest issued by the entity.” That requirement will be removed and the Draft Explanatory Memorandum says this is being done “to capture interest and amounts economically equivalent to interest, in line with the OECD best practice.” In other words, Treasury prefers not to limit the new thin capitalisation measures by tying them to the debt-equity tests.

An entity’s “net debt deductions” is the entity’s “debt deductions” reduced by any amounts of assessable income which are “interest”, “an amount in the nature of interest” and “any other amount that is calculated by reference to the time value of money.”

Tax EBITDA. “Tax EBITDA” is defined in the Exposure Draft quite briefly. In summary, the calculation is:

taxable income for the current year (disregarding the operation of the thin capitalisation rules
plus ‘net debt deductions’ for the income year (ie, interest and similar amounts minus debt deductions)
plus decline in value of depreciating assets calculated under Div 40-B (only)
plus capital works deductions calculated under Div 43
plus tax losses of earlier income years being deducted in calculating this year’s taxable income

Some implications.
This definition has some important implications:

  • the term is “tax EBITDA” – that is, it is defined using tax law concepts rather than the accounting concepts which the label might suggest,
  • only amounts of assessable income count for this purpose; amounts which are NANE (such as dividends from foreign subsidiaries or gains on the sale of shares in active foreign subsidiaries) or exempt income do not count as “Earnings” and so reduce the available envelope,
  • using the item “net debt deductions” is similarly unfriendly to taxpayers. As noted above, the term “debt deductions” is already drawn very broadly and will be expanded further, but the amounts which can offset debt deductions is narrower, and is limited just to “interest,” “an amount in the nature of interest” and “any other amount that is calculated by reference to the time value of money”,
  • to the same effect, adding back as the “Depreciation” component only deductions claimed under Subdiv 40-B and Div 43 means tax EBITDA will still be depleted by the impact of:
    • deductions calculated under Div 328 (which would seem a drafting oversight), or
    • amounts which are deducted under Div 40, but not Subdiv 40-B, such as balancing adjustments, in-house software development pool amounts, project pool amounts, certain primary production amounts, exploration and prospecting, mining site rehabilitation, black holes and so on.

Segregated functions in groups.  It is quite likely that some (non-consolidated) groups will have a group structure where assets are owned by operating entities and debt is raised (and on-lent) by finance entities. Once the group is owned from offshore or has foreign operations, this kind of structure will be problematic under an earnings-based rule unless the finance entity on-lends to the operating entities at interest to ensure that it has only modest “net debt deductions” which might exceed the limit. If the finance entity injects the funds interest-free or as equity into the operating entity, it may face obstacles in being able to satisfy an earnings-based test. The same issues would arise if the finance entity is a trust.

The driver of this problem is that (outside consolidation) the rules do not allow for the aggregation of structures where income is earned by the operating entity but interest is incurred by the finance entity. Thus, a project which might meet the 30% of EBITDA test considered as a whole, may fail because of separation of functions.

15 year carry forward.  If debt deductions are disallowed under the fixed ratio rule, the amount of the denied deductions in a year can be carried forward for up to 15 years (assuming there is ‘headroom’ in the subsequent year to absorb some or all the denied amount). This option is only available for amounts denied under the fixed ratio rule which means it is available only –

  • to general investors (other investors cannot use this rule),
  • who are using the fixed ratio test in the year in which the deduction is denied (rather than the group ratio rule or external third party debt rule), and
  • the taxpayer must be using the fixed ratio rule in the subsequent year and every intervening year as well (effectively removing the ability to make an election if the taxpayer ever hopes to use its denied deductions).

The Draft proposes to deny companies the ability to carry forward denied deductions if there is a change to the majority underlying ownership of the company (determined using the rules which apply for the purposes of loss carry forward). There is no similar business test alternative available to preserve the denied deductions after a change of ownership. The denial is not triggered if there is a change to the ownership of a trust.

There are also complex amendments dealing with the treatment of deferred deductions if a taxpayer with deferred deductions enters a consolidated group. The rules for transferring deferred deductions mirror existing rules dealing with the transfer of tax losses on entry into a consolidated group, but with some variations. For example, denied deductions that are transferred to a consolidated group when an entity joins will reduce the cost which can be pushed down onto the group’s assets but importantly there is no option to cancel the transfer of deferred deductions and leave cost unaffected. It also seems that deferred deductions cannot be transferred to MEC groups.

3.3   Group Ratio Test

The new group ratio test (the group’s third-party interest expense as a percentage of the group’s tax EBITDA) will replace the existing worldwide gearing test (the group’s level of worldwide debt as a percentage of the group’s worldwide equity). The Draft Explanatory Memorandum says, “the group ratio test allows an entity in a highly leveraged group to deduct net debt deductions in excess of the amount permitted under the fixed ratio rule, based on a relevant financial ratio of the worldwide group.”

The group ratio test disallows a portion of the debt deductions of a local entity if the net debt deductions exceed the group ratio earnings limit for the income year. So this test operates using a mixture of financial accounting concepts and Australian tax law concepts:

  • the group is defined to consist of all entities which are consolidated on a line-by-line basis in accounts of a “worldwide parent entity;”
  • the group ratio (the ratio of interest to EBITDA of the group) is calculated using the data from the audited consolidated financial accounts; and
  • that ratio is then applied to the “tax EBITDA” of the Australian entity for the income year; that is, to amounts defined under the Australian tax law.

Group ratio earnings limit.  As mentioned above, the group ratio earnings limit is calculated using information from the consolidated financial accounts of the group. It is the ratio of the “group net third party interest expense” for a period to the “group EBITDA” for the period:

  • the group net third party interest expense is the total amount appearing in financial statements which reflects:
interest expense
plus “amounts in the nature of interest
plus “any other amount that is calculated by reference to the time value of money”
minus payments made by the entity to an (ungrouped) associate of the entity (ie, this amount is not third party interest expense)
minus payments made by an (ungrouped) associate of the entity to the entity (ie, this is not third party interest income and so does not diminish the net interest expense).
  • the group EBITDA for a period is –
the group’s net profit
plus the group’s adjusted net third party interest expense
plus the group’s depreciation and amortisation expenses
minus tax expense

Because this test relies on the group’s consolidated financial statements which might be held offshore or incorporate data from subsidiaries in other countries which is similarly inaccessible to the ATO, special record keeping rules require the entity to keep dedicated records showing how the group earnings ratio limit was calculated.

Implications: Again, these requirements have some important implications:

  • this option is only available to a group with a single worldwide parent entity,
  • the parent entity must prepare consolidated financial statements and have them audited,
  • another requirement insists that the group EBITDA must not be less than zero, so this test cannot be used if the group is in losses worldwide (even if the Australian entity has taxable income).
3.4  External third-party debt test

The existing arm’s length debt test has been repealed for general class investors (and financial investors). Instead, they can choose to use the new, much narrower, external third-party debt test and all the group’s interest expense on qualifying debt will be unaffected by thin capitalisation considerations.

The decision to replace the arm’s length debt test contradicts the ALP’s policy position for the 2022 election that it would limit debt-related deductions using a test based on profits “… while maintaining the arm’s length test…” However, the kind of debt which will satisfy this new test is subject to pre-conditions which indicate the parties were dealing at arm’s length. But this decision, and Treasury’s decided lack of enthusiasm for using arm’s length tests, was strongly hinted at in Treasury’s Consultation Paper (August 2022) – Treasury was very concerned that the impact of the fixed ratio rule might induce taxpayers to start employing the arm’s length debt test instead. Consequently, “arm’s length debt” has been taken off the table and replaced with “third party debt.”

That is, under the new rules deductions may be denied for debt issued on arm’s length terms if it is issued to an associate.

The external third party debt test disallows the entity’s debt deductions for the income year if they exceed the entity’s external third party earnings limit. To put this the other way, an entity can deduct all the interest on debt it borrows from third parties (provided it only uses the funds in Australia to produce assessable income and is only secured over assets held by the entity, subject to the conduit financing rule).

External third party earnings limit. The external third party earnings limit starts from the debt deductions attributable to a “debt interest issued by the entity” (which effectively reintroduces the debt test). The test then removes –

  • any debt interest that was issued to an associate entity of the entity,
  • any debt interest held by an associate entity of the entity at any time during the income year,
  • any debt interest where the holder has recourse to the assets of some other entity, and
  • any debt which is used, even in part, to fund operations outside Australia if they produce NANE or exempt income (such as funding the assets or operations of a foreign PE).

In these rules, the level of ownership which will make an entity an “associate entity” will be reduced from the standard 50% to 10%.

Conduit financing structures.  The regime does contain one concession to the common practice of using a “Fin Co” to raise funds and have it on-lend to other members of the (non-consolidated) group.

The rules allow the on-lending of externally-raised debt to associates of the borrower by relaxing some of the conditions that would ordinarily prohibit this (the associate tests and the recourse test), but the conditions which must be met are very tight, in essence requiring a back-to-back loan arrangement of the borrowed funds (only) and on the same terms. The requirement that the loan arrangements be between “associate entities” may also cause problems for stapled structures.

The initial borrower and the entities to which it lends must all make an irrevocable to apply the regime and security can only be over the assets of the Fin Co and the ultimate borrower(s).

Election requirements: General class investors cannot make this choice if:

  • the entity has one or more associate entities who are general class investors for the income year; and
  • those associate entities are not exempt from the thin capitalisation rules (although the drafting in the proposed legislation implies the opposite); and
  • at least one of the associate entities does not make a choice to use the external third party debt test.

The EM states that “(t)he restriction on this choice ensures that general class investors and their associates are not able to structure their affairs in a way that allows them to artificially maximise their tax benefits by applying a combination of different thin capitalisation tests.” However, with such a low associate entity threshold, there may be instances where entities with a 10% common shareholder are not aware they are associate entities of each other, let alone what choices they may make.

Implications: Again, these requirements have some important implications:

  • the requirement that the holder may only have recourse to the assets of the entity which borrowed the money is unfortunate (and odd). This means debt with a parental guarantee cannot be external third party debt even though it has been borrowed from an unrelated lender (and the main effect of the parental guarantee will be to drive down the borrower’s interest cost – which should be a good thing!),
  • the back-to-back loan regime is potentially helpful (if some of the strict requirements such as identical terms are removed) but it only assists taxpayers where the initial borrower and the entities to which it is lending are all using the external third party debt test,
  • the requirement that all associate entities make the same choice will be problematic for portfolio companies of private capital funds, including both private equity and venture capital, and
  • even though the test is unnecessarily constrained, a taxpayer with only modest amounts of related party debt or non-qualifying debt might decide to apply this test even though it comes at the cost of not deducting the interest on some of its debt.
3.5  Operative rules

Amount denied. The operative rules all deny deductions to a taxpayer once the level of debt is excessive, but they work slightly differently depending on which regime is being applied:

  • if the taxpayer is applying the fixed ratio rule or the group ratio rule, the amount disallowed is the amount by which the entity’s net debt deductions for the income year exceed the entity’s fixed ratio earnings limit. That is, any interest income earned by the taxpayer reduces the amount of deductions potentially being denied, and
  • if the taxpayer is applying the external third party debt test, the amount disallowed is the amount by which the entity’s gross debt deductions for the income year exceed the entity’s external third party earnings limit.

The explanation in the Draft Explanatory Memorandum says this difference is drawn, “to ensure that [the external third party debt test] achieves its policy of effectively denying all debt deductions which are attributable to related party debt.” This statement is curious; the taxpayer could be earning interest from the same external entity that was its financier.

Impact on deductions. Once the amount denied has been calculated, the impact is then spread across all debt deductions incurred by the taxpayer, reducing each debt deduction by the same percentage. The current rules work in a similar way.

4.   Thin capitalisation and transfer pricing

There has always been a complex interplay in Australian tax law between the thin capitalisation and transfer pricing rules.

One issue that has arisen is whether transfer pricing rules could potentially affect the amount of debt taken on by a taxpayer, as well as the price of the debt. Could a taxpayer find itself under challenge from transfer pricing rules for an amount of debt which was within thin capitalisation thresholds?

The argument was apparently put to the ATO that the enactment of thin capitalisation rules, which explicitly referred to the debt:assets test as a “safe harbour,” impliedly limited transfer pricing disputes just to disagreements about the price and not the size of any debt. In 2010 the ATO rejected that argument noting,

… at least since the 1990s consideration of the debt and capital structure has consistently been a consideration in achieving an arm’s length outcome in relation to risk reviews, audits and advance pricing arrangements. In some cases this has been as direct as asking the taxpayer to address the high level of debt by injecting equity …

During the re-writing of Australia’s transfer pricing rules in 2012 and 2013, one of the more incomprehensible provisions in the transfer pricing rules was included to try to sort out the interaction between transfer pricing and thin capitalisation. The drafter of the Explanatory Memorandum to the 2013 Bill said the provision was meant to protect the thin capitalisation rules from interference from the transfer pricing rules: “The rule preserves the role of Division 820 in respect of its application to an entity’s amount of debt [and] ensures that Subdivision 815-B does not prevent the operation of the thin capitalisation rules.” The author of the Explanatory Memorandum to the 2023 Draft takes a similar view: the provision “effectively disapplied [transfer pricing] in relation to the quantum of the debt interest.”

The 2023 Draft deliberately revisits this balance: an amendment to the transfer pricing rules will now re-instate transfer pricing considerations in determining the amount of debt permitted for general class investors if they are using the fixed ratio limit or group ratio limit.

5.   Other matters

The Draft makes some other changes to current law which are worth noting.

Section 25-90.  The decision to repeal s. 25-90 is a major change. It is discussed in detail in our Tax Insight available here.

Financial entity.  The range of entities which can qualify as a ‘financial entity’ for the thin capitalisation rules is being narrowed by removing “an entity other than an ADI that is … a registered corporation under the Financial Sector (Collection of Data) Act 2001 …”  According to the Explanatory Memorandum, many companies “can register under that Act for reasons unrelated to income tax” but “an increasing number of entities are now purporting (for tax purposes) to be financial entities.”  This category is being removed for integrity reasons.

Associates entities. It was noted above that an entity can be affected by the thin capitalisation rules if it is an “associate entity” of an “outward investor.” The Draft notes that this can cause problems for the superannuation sector.

A large retail or industry fund may have sizeable investments in many entities and thus a large number of “associate entities”. The “association” tests work in both directions:

  • if an investee is an outward investor and the fund is an associate entity, the fund becomes an outward investor as well even though all the fund members are residents and the fund has no offshore investments or operations of its own; and
  • if the fund is an outward investor and the investee is an associate entity, the investee becomes an outward investor as well even though the investee is not owned from offshore and may have no offshore operations of its own.

The associate entity tests for these purposes apply a 50% ownership test, rather than the 10% test applied in other parts of the thin capitalisation rules.

While the fund itself is generally prohibited from borrowing (and so has little to fear from the thin capitalisation rules), the associated entities may be adversely affected by the thin capitalisation rules: in other words, a resident company with no foreign operations may find its debt deductions are now in jeopardy because it is highly leveraged.

The Draft proposes to resolve this problem by amending the definition of “associate entity” so that it does not apply to a trustee of a complying superannuation entity (other than an SMSF). The apparent intention is to solve both problems:

  • to immunise a fund with some borrowings from being exposed to the thin capitalisation regime just because it has invested in an outward investor, and
  • to immunise an investee with borrowings from being exposed to the thin capitalisation regime because a retail or industry fund, which is an outward investor, has invested in it.

6.   Dates

The new rules will apply for income years commencing on or after 1 July 2023 with no grandfathering of existing debt or grace period to allow for the reorganisation of existing structures.

Treasury has set a deadline of 13 April 2023 for submissions on the Exposure Draft.

Toby Eggleston
Toby Eggleston
+61 3 9288 1454
Ryan Leslie
Ryan Leslie
+61 3 9288 1411
Nick Heggart
Nick Heggart
+61 8 9211 7593
Graeme Cooper
Graeme Cooper
+61 2 9322 4081

Tax Insight: New limits on deducting interest

Treasury decided to spring a very big surprise on multinationals in an Exposure Draft released on 16 March 2023. The Draft was expected to address the Government’s promise to reconceptualise the thin capitalisation safe harbour from a debt: asset ratio to a percentage of EBITDA, which it does (a detailed analysis will follow). The Draft was not expected to re-write rules about deducting interest incurred to capitalise or buy foreign subsidiaries.


The ability of Australian companies to deduct interest on money borrowed onshore to capitalise their foreign subsidiaries with equity (or buy the shares in a target company) has always been an uneasy combination of 2 issues:

  • whether Australian tax law treated dividends received from the foreign subsidiaries as
    • assessable income, but with a credit for foreign tax (the rule just prior to 1990), or
    • non-assessable and non-exempt income (the rule from 1990), and
  • whether the treatment of the interest expense
    • followed the treatment of the dividend income (deductible if the dividend was assessable, and not deductible if the dividend was NANE) (the rule prior to 2001), or
    • was subject to a special rule to allow the interest deduction even if the dividends were NANE (the rule from 2001).

Those 2 rules have rarely been stable but the most recent changes to them were driven by Treasury pursuing sensible policy considerations:

  • the shift in the treatment of dividends from assessable to NANE in 1990 was made because, “the effect of the [foreign tax credit system] for companies … broadly equivalent to providing an exemption for the dividends, but it imposes greater compliance costs than would an exemption,” and
  • the decision to allow a deduction for interest incurred to earn NANE dividends in 2001 was made because, “debt deductions … will come within the scope of the thin capitalisation regime when determining the amount to be allowed” – in other words, the extension of the thin capitalisation regime to outbound investments was the appropriate method of preventing the Australian operations from carrying excessive levels of debt.

Treasury was happy with that theory in 2001, but apparently there was a contrary view inside the organisation because in May 2013 in the 2013-14 Budget, the Government announced that it would be, “removing the provision allowing a tax deduction for interest expenses incurred in deriving certain exempt foreign income” as part of a suite of measures to “address profit shifting by multinationals through the disproportionate allocation of debt to Australia.” No legislation was enacted before the election, and it seemed the proposal had died:

  • in November 2013, the incoming Government announced that it would not proceed with the May 2013 announcement but would instead “introduce a targeted anti‑avoidance provision after detailed consultation with stakeholders.” This measure too was never enacted,
  • in October 2015, the OECD issued its Final Report on BEPS Action 4 which criticised repeatedly “groups using third party or intra-group financing to fund the generation of exempt income.” Treasurer Scott Morrison did not address this part of the Report in his media release about Australia’s responses to the BEPS project, noting only that “Australia has already tightened its Thin Capitalisation rules,”
  • the Senate Economics Committee inquiry into Corporate Tax Avoidance examined the problem it termed “debt loading” but the Final Report in 2018 did not pursue the idea of repealing s. 25-90, recommending instead repealing the safe harbour and arm’s length debt tests, and using worldwide gearing as the only rule for thin capitalisation purposes, and
  • in the run-up to the 2022 election, the ALP’s tax policies focussed on changing the mechanics of the thin capitalisation safe harbour and a suite of other measures directed to multinationals; there was no mention of any change to s. 25-90.

So, the inclusion of 2 brief paragraphs in the Draft released yesterday, effectively repealing the ability of companies to deduct interest incurred to earn NANE dividends, was truly remarkable. The Explanatory Memorandum said that decision was made because:

 The rules … go against the policy underlying the new rules as it gives rise to a double benefit; the benefit of the income being NANE income and the benefit of a  deduction for the interest expenses incurred to derive such NANE income.

This bland statement doesn’t explain why that very position was acceptable tax policy for the last 20 years, why a thin capitalisation rule based on cash flows requires a different rule to one based on asset values, nor why the newly-tightened thin capitalisation rules will apparently be inadequate to protect the Australian tax system – especially when NANE income is excluded when calculating tax EBITDA for the purposes of the new 30% of EBITDA fixed ratio thin capitalisation test.

The measure

The Draft contains a simple 2-line proposal to remove references to s. 768-5 from s. 25-90 and the equivalent provision in the TOFA regime (s. 230-15). (Interest would still be deductible if incurred in connection with NANE dividends paid out of income already attributed from a CFC or FIF, but these dividends are made NANE because the profits out of which they are paid have already been attributed and taxed).

Assuming it is passed as drafted, the new measure will start for interest incurred in income years starting on and after 1 July 2023. There is currently no provision which would “grandfather” existing debt structures or allow a grace period for taxpayers to unwind them before the change applies.

The future

Assuming it is passed as drafted, we will return to 2000: a world of tracing of loan funds and apportionment for interest expense. One reason for the 2001 amendments was to get rid of the compliance nuisance that “the debts of an outbound investor are traced to an end use to determine the treatment of the interest expense.”

We will also likely return to a world of cash damming, quarantining funds and similar self-help practices to mitigate against the impact of the rule. It would a brave decision to put in place the kind of 2-tier structure struck down by the High Court in the Consolidated Press case in 2001, but no doubt variations will emerge over time.

The Draft is open for submissions until 13 April 2023.

Key contacts

Toby Eggleston
Toby Eggleston
+61 3 9288 1454
Ryan Leslie
Ryan Leslie
+61 3 9288 1411
Nick Heggart
Nick Heggart
+61 8 9211 7593
Graeme Cooper
Graeme Cooper
+61 2 9322 4081

NSW Duty/Land Tax Alert: foreign surcharges do not apply in relation to some countries

By Jinny Chaimungkalanont, Nick Heggart and Jonathan Wu

Update: As of 29 May 2023, Revenue NSW has added India, Japan, Norway and Switzerland to the list of nations that have international tax treaties with the Federal Government which may affect surcharge purchaser duty and surcharge land tax. Further, Revenue NSW has extended the refund period such that refunds may be available in respect of Surcharge (defined below) payments made on or after 1 January 2021 (previously 1 July 2021).

Update: Interestingly, on 15 March 2023, the State Revenue Office of Victoria (SRO) announced that the position in Victoria has not changed, and the SRO will continue to apply the Victorian provisions to all foreign purchasers and absentee owners.

Currently, duty is payable in NSW on the purchase of residential land at rates of up to 7%. Where a foreign person acquires residential land, surcharge purchaser duty will also be payable at 8%. Similarly, surcharge land tax of 4% is payable on top of regular land tax for foreign persons who own residential land in NSW. Continue reading