The Subject to Tax Rule MLC

In September, nine countries signed the Multilateral Convention to Facilitate the Implementation of the Pillar Two Subject to Tax Rule at a ceremony in Paris.  Other countries are expected to follow. Australia observed, but did not sign. Nevertheless, Australia has indicated its support for the Two Pillars project, including the Subject to Tax Rule, and recently reiterated that support in a Media Release issued in September. Just how the Subject to Tax Rule will affect Australian taxpayers may prove surprising.

Background

The ideas behind the Subject to Tax Rule [‘STTR’] were part of the Two Pillar project from its earliest days. The OECD’s, Public Consultation Document – Addressing the Tax Challenges of the Digitalisation of the Economy (February 2019) described, ‘a subject to tax rule that would apply to undertaxed payments that would otherwise be eligible for relief under a double tax treaty.’ The general idea was repeated in Public Consultation Document – Global Anti-Base Erosion Proposal (November 2019) amplified somewhat in the OECD’s Tax Challenges Arising from Digitalisation – Report on Pillar Two Blueprint (October 2020) and was a component of the Statement on a Two-Pillar Solution to Address the Tax Challenges Arising from the Digitalisation of the Economy (July 2021) and signed by Australia in October 2021.

While not expressed in this way, the STTR was widely understood to be part of the price extracted by the developing world for accepting the rest of the Two Pillar project.  The July 2021 Statement put it this way: ‘[Inclusive Framework] members recognise that the STTR is an integral part of achieving a consensus on Pillar Two for developing countries.’ It went on to say, Inclusive Framework members ‘would implement the STTR into their bilateral treaties with developing IF members when requested to do so’. Australia committed to implementing the Two Pillar project by signing the October 2021 Statement which repeated that commitment about the STTR.

The detail of the STTR has evolved a little since 2019, but the core idea is, the STTR allows a source state to reinstate a domestic tax claim if a treaty limits the source state’s right to tax certain kinds of income, paid to an associated entity resident in the other contracting state, but the recipient state does not impose tax on the receipt of at least 9%. Where those conditions are met, certain limitations in the treaty are relaxed to allow the source state’s claim under domestic law to operate, with the tax capped at 9% of the gross payment. Because the outcome would involve imposing tax in a manner inconsistent with a treaty, it was always acknowledged that the STTR required changes to bilateral treaties. The OECD’s drafters offered two routes for amending treaties: a model article and commentary for use by countries that wished to give effect to the STTR by adding a new clause to their existing bilateral treaty, and a multilateral convention [‘MLC’] that would speed up implementation by amending multiple treaties in one instrument. After a lengthy delay, the model Article and Commentary were released in July 2023 and the STTR MLC was released in September 2023 and opened for signature from October 2023. It is this STTR MLC that was signed in Paris last month.

Key design features of the STTR MLC

Scope. The STTR MLC is relevant only where a bilateral treaty affects the ability of the source state to impose tax on income sourced in the state and earned by a resident of the other contracting state.

The scope is further limited to exclude amounts earned by a recipient which is an individual, a non-profit organisation, the other contracting state entity, or an international organisation. Further provisions can exclude amounts earned by certain investment funds, pension funds and entities which are formed to hold assets for them.

The payer and the recipient must be connected entities, which is set at 50% common ownership, and buttressed by an anti-avoidance rule if there are unrelated interposed entities.

In addition, there is a materiality threshold: the STTR MLC only applies if the covered income for the year earned by connected entities in the other contracting state exceeds €1m (or €250,000 for jurisdictions with GDP below €40 bn).

And there is a profitability threshold: the STTR MLC only applies if the amount of covered income for the year earned by connected entities in the other contracting state exceeds the costs of generating that income by 8.5% or more.  (This profitability threshold does not apply to interest or royalties.)

Covered income.  The STTR MLC only applies to ‘Covered Income’ a defined term which means:

• interest, guarantee fees or ‘financing fees’,
• royalties and payments for distribution rights for a product or service,
• premiums for insurance or reinsurance,
• rental payments for use of industrial, commercial or scientific equipment, or
• payments for services.

Dividends are not on the list, so the fact that the residence state exempts dividends from foreign subsidiaries is irrelevant.  Similarly, capital gains are not on the list, so the fact that the residence state exempts capital gains is irrelevant.  This may make sense if the residence state were exempting capital gains made on the sale of shares in a subsidiary, as it would be consistent with the omission of dividends, but it is not obvious why capital gains are omitted if the residence state is exempting capital gains in other circumstances.

Another exception excludes certain shipping income.

Limiting treaty provisions. A further precondition requires that ‘the tax that may be charged in a contracting jurisdiction … is limited …’, and is limited by the articles dealing with business profits (art 7), interest (art 11), royalties (art 12) or other income (art 21).  The source state’s tax on interest and royalty income is sufficiently ‘limited’ if the source state is prohibited from taxing the amounts at all, or if there is a limit on the rate at which the source state may tax the amount.

To put this the other way, the STTR MLC is not enlivened if the source state’s tax claim is limited by other provisions of the treaty. 

This is not as surprising as it may seem. Some articles which limit source state tax (such as articles 15, 18, 19 and 20) are relevant for payments to non-resident individuals, but all payments to individuals are already excluded from the scope of the MLC so omitting any reference to these articles is probably not material.  Similarly, the limit in art 8 on taxing income derived from using ships or aircraft in international traffic is immaterial for shipping income (because it is not covered income) though it will matter for income from aircraft.

But the omission of capital gains made (art 13) is worth noting.  The Public Consultation Document in February 2019 had proposed that source countries would be able to insist on tax with respect to interest, royalties and capital gains (arts 11-13), but capital gains did not make it into the final list in the STTR MLC.  This means, if the source state’s domestic tax claim is limited by art 13, and the residence state does not tax the gain sufficiently or at all, the residence cannot insist on its tax by virtue of the STTR MLC.

Residence state tax.  The next condition for triggering the STTR MLC is, the tax rate in the recipient state must be less than 9% of the gross payment.

Ordinarily, the relevant tax rate will be the headline rate of tax of the recipient state: ‘the statutory rate of tax applicable in that jurisdiction on such income …’.  However, the rules say that the tax rate will be the rate applied to the item of income if the item enjoys a ‘preferential adjustment,’ defined to mean a permanent exclusion from income, deduction or tax credit ‘linked to the item of covered income.’ 

Importantly, the tax rate is not the effective rate suffered by the recipient in the residence state, and so deductions and tax credits will not be relevant to computing the 9% tax rate (unless they amount to a ‘preferential adjustment’.)

Effects

In source state. Where the STTR MLC is enlivened, the operative provision of the MLC says, ‘an item of covered income arising in that jurisdiction … may … be taxed in that jurisdiction …’ However, the tax that the source state can collect ‘shall not exceed … the difference between 9% and the tax rate … on that item of covered income …’ imposed in the residence state.

The rate cap means it is also a condition for triggering the STTR MLC that the tax rate in the source state is not already 9% of the gross payment. The STTR MLC makes this explicit by insisting that the operative provisions are not enlivened if the income may be taxed in the source state at, ‘a rate equal to or greater than the specified rate …’.

So the combined effect of these rules is that the space opened up for the source state to reinstate its tax claim is effectively:

               9% rate

less        tax currently imposed in the source state

less        tax imposed in the residence state

The practical effect of this regime will be to mitigate much of the impact the STTR MLC might have had so far as interest and royalties are concerned. Many countries, especially developing countries, will retain in their treaties the right to tax interest and royalties sourced in their country at rates higher than 9%. If that is so, the (say) 10% rate permitted under the bilateral treaty is already higher than that the capped 9% rate which the STTR MLC allows.

In residence state.  In addition, to reinstating the tax claim of the source state, the STTR MLC also has an impact in the residence state.  The MLC specifically provides that the STTR ‘shall not create any obligation under provisions of this agreement’ for the residence state to provide a credit for any tax collected pursuant to the STTR or to exempt income which has been taxed in the source state under the STTR, despite the requirements in art 23 of the treaty.

The drafting is not a prohibition on granting a credit or exemption; rather it offers an option: any tax collected in the source state pursuant to the STTR need not be credited in the residence state. Presumably this means the residence state will have to assess whether a credit or exemption would be triggered under domestic law, and where it would, then decide whether it wants to amend its domestic law to deny that credit and so create double (modest) tax.

Enlivening the MLC

As was noted above, the STTR MLC is one of two pathways to including a STTR in a bilateral tax treaty. At first glance it is obviously reminiscent of the 2017 Multilateral Convention to Implement Tax Treaty Related Measures to Prevent Base Erosion and Profit Shifting [‘MLI’] which amended many of Australia’s bilateral treaties to give effect to recommendations arising from the first BEPS project (2013-15). However, the mechanics of the STTR MLC are somewhat different to the MLI.

The STTR MLC will be enlivened where both parties to a bilateral tax treaty have signed the STTR MLC, and both parties have nominated that treaty as a ‘Covered Tax Agreement.’ However, unlike the MLI, the STTR MLC does not operate as a stand-alone treaty which changes the text of individual articles in a bilateral treaty. Instead, the STTR MLC operates as part of the bilateral treaty. The MLC provides, ‘Annex I (The subject to tax rule) shall be included in all Covered Tax Agreements and shall form an integral part thereof.’  The Commentary to the STTR MLC says, adding this Annex to existing treaties means the STTR MCL ‘functions like an amending protocol to a single existing bilateral tax treaty.’

Secondly, the MLI was notoriously rife with options and elections – it was not sufficient that the two countries both nominated their treaty, they then had to agree which articles of the MIL would amend their bilateral treaty, and they then had to agree which options within that article they preferred! Finding matches to the various combinations has proved a nightmare. (This explains the ATO’s reluctance to set out what our affected treaties now say: the ATO prefaces synthesised texts of treaties with the caution that the synthesised text, ‘is to facilitate the understanding of the application of the MLI to the Agreement’; it may or may not be accurate; the most you can hope for is, if the ATO’s version ‘turns out to be incorrect … the ATO will take that into account’ when deciding penalties!)

The OECD has apparently learned from this debacle and there are only a few options within the STTR MLC, the three most important being:

• the parties must nominate any gross-based taxes they want the STTR MLI to affect,
• whether they want the MLC to include the definition of ‘recognised pension fund’ in their bilateral treaty, and
• whether they want to include the so-called ‘circuit-breaker provision’ (which deals with how to apply the materiality threshold to a country which once met the test for being a developing country but no longer does).

The Australia government has indicated its support for Pillar Two and the STTR but did not sign the STTR MLC in September and so is not (yet) affected by it.  Of the 9 countries which did sign, Australia has a bilateral with three – Indonesia, Romania and Türkiye – but none of these countries nominated their treaty with Australia as a covered tax treaty.  This may indicate that Australia will express its support for the STTR through amended bilateral treaties, but time will tell.

Implications for Australian taxpayers

The effect of the STTR MLI is to open some space for greater source state tax where the preconditions are met. How will that outcome play out for Australian groups?

Inbound income.  As noted above, while the Australia government has indicated its support for the STTR, it has not indicated whether that support will be reflected in bilateral amendments to individual treaties, or by signing the STTR MLI.  Until Australia signs the STTR MLC, or amends a bilateral treaty, source countries will not have any greater taxing rights under the STTR.

But even when Australia does act, it seems unlikely, Australian groups will suffer increased levels of source state tax for income that is remitted to Australia for the obvious reason that there are few categories of ‘covered income’ that would not be fully taxed on repatriation to Australia:

• dividends from foreign subsidiaries which would be NANE in the hands of the Australian parent ought not to be a matter for concern because dividends are not covered income, and may well be subject to withholding tax at source at a rate above 9% already,
• capital gains made on the sale of foreign subsidiaries which would be NANE in the hands of the Australian parent ought not to be a matter for concern because capital gains are not covered income,
• business profits earned through foreign PEs which would be NANE in the hands of the Australian taxpayer ought not to be a matter for concern because of the interplay between the treaty and s. 23AH ITAA 1936. That is, the source state’s ability to tax business profits is only limited in those cases where the non-resident does not have a PE in the state; but in that situation, it is unlikely that the Australian owner would qualify for NANE treatment under s. 23AH, so the pre-conditions for triggering the STTR are unlikely to be met,
• interest and royalties from foreign subsidiaries ought not be a matter for concern because they are most likely fully taxable in Australia, and may well be subject to withholding tax at source at a rate above 9% already.

One scenario where this might be an issue is conduit foreign income [‘CFI’]. This type of income is made NANE under the ITAA 1997. So, while the headline corporate tax rate in Australia is 30%, CFI might be treated as a ‘preferential adjustment’ and so amenable to higher levels of source state tax. If the source country is constrained by the relevant clause of the treaty not to tax the Australian entity, the STTR may be in play to increase the level of source country tax and the overall tax burden.

Transactions occurring offshore.  The most likely scenario where Australian groups will have to consider the effects of the STTR MLC will arise where there are transactions between foreign subsidiaries. While the Australia government has not signed the STTR MLC, 9 countries have signed so far, and a further ten have indicated they will do so.  The 9 countries have nominated various treaties and so it is likely that payments between two foreign subsidiaries will be in play.

It is quite plausible that amounts flowing between foreign subsidiaries which are covered income and enjoy reduced rates in the source state are being earned by entities in low-tax jurisdictions, being jurisdictions which (for some reason) the source state has negotiated a treaty. In that case, the source state may find its domestic taxing power reinstated.

If it were not for the provisions dealing with tax credits, the result for the group might be modest – any greater tax collected in the source state might simply give rise to a larger tax credit or exemption in the residence state. But the result would not be a wash if the credit can’t be used in the residence state and the STTR MLC leaves open the option for the residence state to deny the credit anyway which would mean double taxation.

Outbound income.  Finally, it is worth noting that, even though the STTR is thought of as a measure which benefits developing countries, there is no pre-condition to this effect and the provisions operate reciprocally. This means that Australia could find itself empowered to collect greater amounts tax on amounts paid to related non-residents where the non-resident is not subject to Australian income tax or withholding tax because of limits imposed by articles 7, 11, 12 or 21 of the relevant treaty, and is not fully taxable in its own country for reasons of its domestic law.

It is worth pausing to note, a tax exemption delivered by domestic law such as s. 128F for interest, s. 128B for franked dividends or Div 802 for CFI is not sufficient to trigger the STTR MLC; it must be the case that Australia is not imposing tax on the non-resident because of the treaty.

Assuming the materiality and profitability thresholds are met, there are a few scenarios where this combination might occur.  Consider, for example, interest paid to a foreign financial institution which is resident in the US or UK which is protected from Australian withholding tax under art 11(5). Australia is not taxing the interest payment because of the treaty and if it were the case that the financial institution was not taxable in the US or UK, Australian domestic law would operate unimpeded. The financial institution would have to rely on s. 128F to achieve the current outcome, but it may not be available (and this could become the Australian payer’s problem if there is a gross-up clause in the loan agreement).  Whether this problem turns out to be the case will depend on the rate being imposed on the financial institution (or item of income) in the US or UK.

Another example might arise in the case of payments to hubs in Singapore for procurement or marketing services.  If we assume the Singapore-based supplier has no PE in Australia, our source-based claim to tax the payment for services is potentially excluded by the treaty.  Singapore might not tax the hub (or not tax it fully) depending on whether the amount is received in Singapore. This would seem to meet the pre-conditions for reinstating the Australian tax claim. One solution to this example is that our treaty with Singapore does not remove Australia’s right to tax income which is not remitted to Singapore. Whether marketing and procurement hubs survive the STTR will likely turn on whether the limit on Australian tax in art 7 is enlivened, and the rate being imposed by Singapore on the hub.