Five Curiosities in Australia’s GloBE Regime

by Graeme Cooper, Toby Eggleston, Ryan Leslie, James Pettigrew, Nick Heggart, Jinny Chaimungkalanont

Australia took the next step along the path to implementing the 15% global minimum tax in July with the introduction of three Bills to legislate the framework for the new tax regime.  The project comprises several parts but the component being addressed in the three Bills is usually referred to the Global Base Erosion rules [‘GloBE’], comprising the Income Inclusion Rule [‘IIR’], the Under-Taxed Profits Rule [‘UTPR’] and the Domestic Minimum Top-up Tax [‘DMTT’] measure. They operate in sequence to produce the 15% global minimum tax.  This Tax Insight looks at a few odd elements of Australia’s GloBE rules. There is probably not much Australian legislators and policy-makers can do about oddities embedded in the OECD Model Rules, but Parliament definitely has the power to remedy the new and idiosyncratic problems being introduced by these Bills.

Curiosity No 1: Their GloBE isn’t really tax

One complication of Australia’s own making is how Australian tax law will view tax paid to foreign Governments under their GloBE measures.  The general posture underlying the Bills is, we won’t respect foreign GloBE as tax. The treatment is more nuanced than this because foreign countries may be collecting foreign tax under one or more of the three regimes – a DMTT, an IIR, or an UTPR – and the Bills do respect foreign tax sometimes, but there is no consistent pattern to the places where foreign GloBE is respected or dismissed.

The starkest example of this is in the anti-hybrid rules.  In general terms, Div 832 ITAA 1997 will penalise an Australian payer if it makes a deductible payment to a related foreign recipient which is not taxable on the receipt. As currently drafted, because GloBE isn’t tax, if the foreign country imposes $1 of corporate tax, the Australian company is safe, but if the foreign country imposes $15 of GloBE tax, the Australian company is penalised. This is illogical.

Another example is the FITO rules. As currently drafted, the Bills have the effect that if the foreign country imposes the GloBE rules as a floor to its corporate tax, we will grant a FITO, but if the foreign country imposes GloBE by a separate tax, access to a FITO is limited.  This treatment will follow from amendments to Div 770 ITAA 1997. In order to ensure this limit in domestic law can operate unimpeded by our treaty obligations, a treaty-override has been included in the Bills. It seems likely that foreign GloBE taxes are a ‘Covered tax’ for our bilateral income tax treaties either as a foreign ‘income tax’ or as a tax that is ‘substantially similar [and] imposed after the date of signature of the Convention in addition to … the existing taxes.’ If that is so, Australia would be obliged to give a FITO for foreign GloBE taxes, despite the amendments to Div 770. So, in order to ensure that doesn’t happen, the Consequential Bill amends the International Tax Agreements Act 1953 to override the treaty obligation that would otherwise apply to give a FITO for tax paid under a foreign IIR or UTPR, though Australian residents may still qualify for an income tax FITO for tax paid under a foreign DMTT.

Curiosity No 2: Joint and several liability (and no clear exit)

The Consequential Bill will insert Div 128 into Schedule 1 of the Tax Administration Act. Proposed s. 128-5 provides, ‘if an amount is payable under the Minimum Tax law by a Group Entity … that Group Entity and each other Group Entity … are jointly and severally liable to pay the amount.’

Australia went down this same path twenty years ago when designing the tax consolidation regime.  Treasury’s initial proposal was for all members of a tax consolidated group to face joint and several liability for the unpaid tax debts of any group member. In fact, the proposal now in the Consequential Bill is even more ambitious. At least under the tax consolidation proposal, an Australian seller could sell a foreign subsidiary without that buyer needing to do due diligence on the tax debts of the Australian group: the foreign target could not be a member of the Australian tax consolidated group and so would not inherit the Australian tax debts of the Australian companies. But in this Bill, foreign companies are exposed to the Australian tax debts of the entire accounting group. Buyers of foreign subsidiaries will have to do due diligence on the Australian tax debts of every other member of the Australian group.

The tax consolidation proposal had to be abandoned because of its deleterious impact on commercial activity: a bidder could not buy a particular company and a lender could not lend to a particular company without the bidder or lender needing to conduct due diligence on the entire consolidated group.  No-one could be confident when buying a subsidiary or lending to it, that they weren’t going to be affected by an income tax liability quite disproportionate to the size of the target. A similar solution is needed for unpaid GloBE.  The simplest solution would be to leave the GloBE liability with each entity, but if joint and several liability is to remain, a solution similar to that found in Div 721 ITAA 1997 will have to be created to allow a ‘clear GloBE exit’ for entities leaving MNE groups.

Curiosity No 3: Securitisation vehicles may still be collateral damage

One particular manifestation of this problem arises for securitisation vehicles – that is, vehicles used to give outside investors recourse to a specific asset pool or stream of cashflows. The securitisation industry is predicated on securitisation trusts being ‘income tax neutral’ in the sense that the income tax obligations are the responsibility of the lenders and other investors rather than the vehicle. But in the GloBE realm, it will often be the case that the securitisation vehicle will be consolidated into the financial accounts of the originator (on the basis of ownership or control) so they can’t be ignored.  And it may be the case that the vehicle has profits for GloBE purposes (but no income tax liability) because the tax base or attribution rules for GloBE differ from the rules in the income tax so there is potential GloBE tax liability.

The OECD is alive to the need for specific provisions to accommodate securitisation vehicles and Chapter 6 of the Administrative Guidance released in June 2024 is dedicated to finding solutions. It proposes three escape routes to solve the GloBE problems of securitisation vehicles:

•  IIR: in most cases, a securitisation vehicle will not be a UPE (or even an IPE) within a MNE Group, and consequently, and so the OECD expects it should not be exposed to top-up tax under the IIR;
• UTPR: the Guidance notes that the Model Rules do not dictate just which entities in a country are liable to pay any UTPR and so the Guidance suggests, ‘jurisdictions may exclude Securitisation Entities from liability to top-up taxes under the UTPR’;
• DMTT: the ‘fix’ for the DMTT will happen by a change to the Commentary to the Model Rules: ‘the Inclusive Framework agrees that jurisdictions adopting QDMTTs are not required to impose top-up tax liabilities on SPVs used in securitisation transactions [and] the Commentary will consequently be revised to clarify that a QDMTT liability in respect of a Securitisation Entity should generally be imposed on other Constituent Entities located in the jurisdiction.’

Chapter 6 of the Administrative Guidance also proposes two escape routes to solve the GloBE problems for the rest of the MNE Group caused by securitisation vehicles.  First, ‘the Commentary will also clarify that a … Securitisation Entity [need not be] a Constituent Entity for the purposes of that QDMTT.’ Secondly, there will be more work on how to handle the ETR calculations for securitisation vehicles:                            

The Inclusive Framework … will … consider issuing further Administrative Guidance to ensure that the use of securitisation transactions (and arrangements of the kind entered into by SPVs that give rise to fair value movements) do not result in MNE Groups paying top-up taxes that are not commensurate with the economic profit that the SPV has made from the activities.

The initial draft of Australia’s proposed joint and several liability rules for GloBE taxes would have turned this effort on its head. While the OECD is working to ensure that a liability ‘in respect of a Securitisation Entity should generally be imposed on other Constituent Entities …’, Australia’s rule would have ensured that a liability of other entities could be imposed on the securitisation vehicle. 

The issue was flagged with both the ATO and the Senate Committee examining the Bills and so a subsequent amendment to the initial Bills was tabled which provides that the joint and several liability provisions do not apply to ‘a GloBE Securitisation Entity’ (a term to be defined later in the Rules).

All of that looks very promising – securitisation vehicles should not be exposed to top-up taxes under any of the 3 regimes and should not cause inappropriate outcomes for other group members – but of course, things are never this easy.

Instead, giving effect to this policy will present the first major test of the willingness of Australia’s drafters to depart from the text of the OECD model rules to reflect Australian conditions. The terms used in the OECD drafting are not a good description of Australian industry practice and so there may yet be an impasse: either we depart from the OECD drafting, or Australian industry practice will have to change, or we are at a stalemate and the joint and several liability issue of securitisation vehicles will remain.

Curiosity No 4: Pushing GloBE liabilities onto shareholders (and their shareholders and subsidiaries?)

The problem caused by insisting on joint and several liability is further exacerbated because vicarious liability is made to extend beyond the members of the accounting group.

Proposed s. 128-10 imposes joint and several liability in the case of a GloBE Joint Venture – the label used to describe an entity in which an MNE Group holds at least 50%. In the case of any GloBE liabilities of a GloBE Joint Venture (and its subsidiaries), the joint and several liability extends to its shareholders – ‘a Group Entity of the Applicable MNE Group that holds a Direct Ownership Interest … in the GloBE Joint Venture.’  In other words, if 2 groups each have a 50% stake in a company, then the 2 shareholders are each liable for the GloBE tax debts of the jointly held company and its subsidiaries.

This raises a further dilemma. If a shareholder in the Joint Venture is liable under s. 128-10 for the Australian GloBE debts of the Joint Venture company, is that shareholder’s liability one to which s. 128-5 then applies? It certainly looks like it is: the s. 128-10 liability looks very much like, ‘an amount … payable under the Minimum Tax law by a Group Entity of an Applicable MNE Group for a Fiscal Year.’  If the process is iterative in this way, then all the other members of the shareholder’s accounting group also become liable for the tax debts of the Joint Venture company. This doesn’t appear to be the intention, but it is a plausible interpretation of the drafting.

Curiosity No 5: Australia’s GIR requirements are excessive

Finally, our policy-makers have decided to increase the deadweight cost of GloBE compliance by imposing extra Australia-specific compliance burdens on affected groups. 

The Model Rules agreed by the Inclusive Framework start with the assumption that every entity must complete a GloBE Information Return (‘GIR’) and lodge it with its local tax authorities (article 8.1.1). However, local entities are relieved of this obligation if –

• the GIR is filed by someone elsewhere in the group (usually the Ultimate Parent Entity or a Designated Filing Entity);
• there is a ‘Qualifying Competent Authority Agreement’ between the two countries (which might be a bilateral income tax treaty, a TIEA, the Convention on Mutual Administrative Assistance or perhaps a bespoke information exchange treaty); and
• the local entity tells its tax authorities that this happened and where the GIR was lodged (articles 8.1.2 and 8.1.3).

So the design is, single filing in one country, and thereafter, it is the responsibility of the tax authorities to share the data. The Commentary to the Model Rules is very clear:

The information exchange mechanism … means that many, and perhaps most, tax administrations will receive the GloBE Information Return through an exchange of information mechanism …

Secondly, the system strikes a balance between disclosure and secrecy of tax information by insisting (i) only one jurisdiction gets the entire GIR, (ii) all affected jurisdictions get the portion of the GIR which contains the ‘general information and the corporate structure,’ and (iii) each jurisdiction with taxing rights gets the portion of the GIR which identifies the ‘ETR and Top-up Tax computation, allocation and attribution’ to that jurisdiction.

Australia, however, wants the option to have it all and insist it be provided directly by the local entity.  Proposed s. 127-20(4) allows the ATO to insist that the local entity provide a copy of the worldwide GIR direct to the ATO (and to penalise the local entity if it doesn’t comply) even though the complete GIR has been provided to a foreign government. It is not clear whether this provision is proposed because the ATO doesn’t trust other Governments to share information, or the ATO thinks the negotiated process is just an unnecessary annoyance, or because the ATO wants to get access to the worldwide information, or all three.

Finally …

It seems unlikely that any of the existing problems will be remedied. The inquiry into the Bills by the Senate Economics Legislation Committee has already concluded. The Committee’s report recommended succinctly, ‘that the bills be passed.’

Other idiosyncratic quirks will undoubtedly emerge as taxpayers and advisors become more familiar with the local provisions. Perhaps future quirks will be remedied, even if the current batch are not.