Our Tax Note last week alerted readers to the release of Treasury’s Discussion Paper seeking submissions on amendments to Australia’s international tax regime and current tax disclosure rules. This Tax Insight examines the Paper’s international tax measures in more detail and reflects on what the final regime might look like and what aspects might change during the consultations.
Thin capitalisation rules became a formal feature of Australia’s tax legislation in 1987 and have been revised on several occasions as governments tinker with the policy settings. The original rules focussed just on possible “earnings stripping”: interest paid to non-resident entities, related to the payer, and tested using a ratio of debt to equity. The major revisions in 2001 expanded the focus to examine capital structure more generally: the rules would extend to interest incurred in Australia in respect of offshore operations, paid to related and unrelated entities, and tested using a ratio of debt to assets (with the possibility of using alternate tests based on world-wide gearing or an “arm’s length” level of debt).
The OECD’s report on BEPS Action 4, released in 2015, included a recommended “common approach” to setting limits on interest deductions, using a test based on the ratio of interest to earnings. The Coalition Government decided not to pursue this approach, saying instead, “Australia has already tightened its thin capitalisation rules,” a statement which was correct, but rather side-stepped the recommendation. The ALP policy for the 2022 election was to shift to the OECD’s common approach.
The Paper seeks answers to a series of questions, some of which are a little curious (“would introducing a fixed ratio rule encourage entities … to shift to an arm’s length test”), some of which should already be known by the government (“what types of entities currently use the existing worldwide group test”) while others seem to be wishful thinking (“would additional limitations be required to prevent any unintended consequences …”, “are there any other changes … that could be made to the arm’s length debt test, to keep in line with the Government’s commitment to limit interest deductions”).
There are hints and innuendoes in the Paper that the final policy could look something like this:
- a de minimis threshold is likely to remain;
- any new test will presumably still be applied to the Australian tax consolidated group as a single entity;
- the new rule will still apply to interest paid to related and unrelated entities, and to onshore and offshore recipients;
- the new rule will continue to apply to interest and a broader group of debt-related deductions;
- the new earnings-based test will apply only to general commercial entities; ADIs and financial institutions will continue to apply the current assets-based test;
- there is the possibility that the current worldwide-gearing ratio will be replaced by the OECD variant, “a group ratio rule” as a means of accommodating groups that are highly-geared worldwide. The OECD version disallows a deduction only if the domestic interest expense under scrutiny exceeds both the earnings-based test on a (domestic group) stand-alone basis, and the ratio of net external debt to worldwide earnings;
- interestingly, the OECD group ratio regime contemplates an uplift of 10% on the actual ratio of worldwide debt – if the group’s external interest is (say), 40% of group EBITDA, the local entity’s deduction might be permitted up to 44%;
- there is a concern that tightening the earning-based rule will simply encourage taxpayers to switch to the arm’s length debt test. Treasury and the ATO are clearly alarmed at this prospect and there is a suggestion that the arm’s length debt test might be withdrawn or curtailed (“[is] the existing legislative architecture for the arm’s length debt test … still appropriate”);
- the limitation will only apply to a group’s “net interest expense.” This should mean that a group which borrows externally through a single financing entity will not be affected, provided the borrower on-lends the borrowed funds at interest within the group;
- there appears to be some appetite for allowing higher gearing of “assets/projects that provide net public benefits or are considered nationally significant…”
But the Paper does not explicitly consider some other important design questions. The most important is, can denied interest be used in other years? It is already well understood that the shift to an earnings-based test will penalise groups that are asset-rich but cash-flow constrained in a particular year – they might be Australian subsidiaries of start-up entities, entities undergoing a restructure, entities owned by private equity investors, entities affected by a temporary price slump, entities investing in projects with a long lead time, entities in the infrastructure sector and so on. The current asset-based safe harbour is much less susceptible to being triggered unexpectedly as asset values tend to be more robust and stable than revenue flows.
It is for this reason the report on BEPS Action 4 specifically provides for “the carry forward of disallowed interest expense and/or unused interest capacity … for use in future years [to] reduce the impact of earnings volatility …” The US rule [Internal Revenue Code, s. 163(j)] allows companies to carry forward excess interest deductions indefinitely. The EU Council Directive (EU) 2016/1164 which mandates the interest limitation for Member States allows them to offer a carry-forward option and even a carry-back option; many Members States offer a carry forward, though sometimes with a time limit.
Nor is there any discussion in the Paper of the interaction between the deduction denial rule and interest withholding tax. It seems quite likely that IWT will have been remitted by the local payer on interest paid during the year, only to discover at the end of the year that some of the interest is non-deductible. Perhaps this outcome is intended, but if not, some adjustment or refund mechanism is needed to compensate the non-resident. Similar (but worse) outcomes arise where the interest for which deductions are denied is paid to an Australian lender paying income tax at Australian marginal rates, which (absent exceptions or reliance on the arm’s length test) is likely to be the case for Australian institutional investors who invest alongside non-residents in private consortia.
Finally, the Paper does not examine issues of transition: when does the new rule start and how will it deal with funding arrangements already on foot? One might expect there will need to be some grandfathering or perhaps phased-in implementation so as not affect existing investments too adversely.
There will be significant pressure for the new regime to address these omissions.
The treatment of royalties has recently received a lot of attention from our tax authorities both to decrease Australia’s tax on royalties (the former government announced a Patent Box regime in its last Budget, and changes are proposed to adjust Australia’s treaty with India to reduce Australia’s claim to tax royalties) and to increase the tax (the ATO released 3 Taxpayer Alerts in the last 4 years involving royalties and outlining scenarios it regards as abusive).
And now the Treasury Paper has outlined a new proposal to deny deductions to resident taxpayers paying royalties to non-residents which are not subject to sufficient tax (typically because they are located in a country with a patent box regime).
The Paper acknowledges that Australia already has a variety of anti-abuse rules including transfer pricing rules, Part IVA, rules which limit access to reduced withholding tax rates benefits in individual treaties, the Principal Purposes Test in the Multilateral Instrument and the possibility of repatriating royalties paid to offshore subsidiaries under Australia’s CFC regime. Nevertheless, the Paper says, “Australia’s tax framework needs a specific measure targeting integrity issues associated with intangibles and royalties.” The Discussion Paper seeks submission on 4 key aspects of the design of this new rule: should the new regime apply to:
- cross-border payments made by all multi-national enterprises, or only those which are also significant global entities;
- payments which are “royalties” under current definitions, or also royalty-like payments which are “embedded” in bigger payments for services, part-sales or other broader bundles of rights;
- payments to all offshore entities or only to related entities; and
- what test should be used to decide whether the recipient is subject to sufficient tax?
But it will be hard to answer those questions without a clearer indication of the concern (or concerns) which is to be the focus for attention. The Paper presents an undifferentiated collection of items:
- the one-time transfer of intangibles developed in Australia;
- the ongoing consequences of the emigration of intangibles developed in Australia;
- payments for the use of intangibles developed offshore:
- the problem of royalties embedded in non-royalty payments made for a broader bundle of rights;
- the problem of non-royalty payments leaving Australia but funding a royalty made by an entity higher up the corporate chain and
- permutations where the transactions do (or do not) involve associates, and where the associates are (or aren’t) located in a jurisdiction with a Patent Box regime.
We know that the project will lead to a rule which denies a deduction to the resident payer, which suggests the rule will not be directed at the transfer of intangibles offshore – there is unlikely to be a deduction to be denied (unless Treasury is thinking about retrospectively denying deductions for the development of any intangible which leaves Australia’s shores, which does not appear to be the case).
But beyond that, there are hints Treasury wants the regime to be very broadly drawn: extending to almost all business payers, to payments which are not currently viewed as royalties, for intangibles developed anywhere in the world, paid to entities whether associates or not, and even if recipients are resident in countries with comparable tax systems.
The discussion paper refers to these proposals as a means to “level the playing field for Australian businesses”. However, given most royalty payments will already be subject to withholding tax this proposal has the effect of actively tilting the playing field it in favour of an Australian based company and undermines the integrity of the negotiated outcomes under Australia’s double treaty network.
As with the thin capitalisation section, there are important design matters which have not been addressed such as –
- there is no discussion about the interaction of the rule denying deductions to payers with the liability of recipients to Australian royalty withholding tax;
- while the tenor of the discussion is about “arrangements … to shift profits from higher tax jurisdictions to low or no tax jurisdictions …” the Paper does not propose any element of “purpose” or “design” or “intention” as a pre-condition to triggering the rule. The so-called “integrity rule” for payments of interest to low-taxed entities offshore, which may well be a model for this regime, specifically includes a requirement that “the entity … entered into … the scheme … for a principal purpose of …” securing the deduction;
- while the Paper specifically mentions emigration of intangibles as abusive (“… situations [where] Australian taxable profits have been reduced due to reduced worldwide income being derived from the exploitation of intangibles in Australia as the intangible is not in Australia …”) the Paper offers no indication about how that situation might be addressed; and
- the interaction of these rules in the event Pillar 2 is enacted, where 15% effective tax would represent an acceptable base level of taxation.
For further information, please contact:
Toby Eggleston, Partner, Herbert Smith Freehills