Minimum tax on discretionary trusts: Definitional challenges would broaden impact

A new proposed minimum tax on discretionary trusts, aimed at preventing income-splitting, could have unintended consequences for many other types of trust arrangements if the federal government does not appropriately limit the scope of that minimum tax. For example, property and infrastructure groups with multi-layered structures may fall under proposed definitions of ‘discretionary trust’. Trustees should be aware of these proposed rules and plan accordingly. 

Key takeaways for decision-makers

• Treasury has proposed a 30% minimum tax on discretionary trusts, effective from 1 July 2028. Its commentary focusses on deterring income-splitting, which is the practice of allocating taxable income to beneficiaries with lower tax rates. This suggests a measure targeted at discretionary trusts that operate in the context of close groups of individuals, which usually involves distinct features relating to allocation of taxable income by a trustee across that group (typically a family group).
   
• However, trusts appear in many commercial contexts in Australia. The measure risks impacting arrangements that are not traditionally the subject of income-splitting concerns. For property and infrastructure groups with multi-layered structures, a ‘look-through’ or group-based approach would be essential to avoid double taxation and disproportionate compliance costs in structures with no income-splitting function.
   
• Part of the challenge is that there is no universally accepted legal or tax definition of ‘discretionary trust’ in Australian law. The definitional architecture chosen will determine whether the measure achieves its stated objective, or inadvertently captures broad swathes of commercial trust structures.
   
• The existing fixed trust rules in Schedule 2F of the Income Tax Assessment Act 1936 (Cth) (ITAA 1936) impose a ‘vested and indefeasible’ standard that is too strict for this purpose. Many commercial trusts with genuinely fixed entitlements, including property and infrastructure unit trusts, would fail due to routine administrative discretions. A new definition focused on the substantive ability to allocate taxable income non-proportionally, rather than the mere existence of any defeasibility, would better target the identified issue.
   
• Despite its prevalence, the concept of a ‘discretionary trust’ has never had a single, settled normative definition in Australian tax law. The existing income tax framework approaches the concept negatively, defining it, in effect, as any trust that is not a ‘fixed trust’ within the meaning of section 272-5 of Schedule 2F to the ITAA 1936. This negative definition has sufficed for the trust loss provisions where it originated, but it was never designed to serve as the gateway to a broad-based minimum tax.
   
• If the current fixed trust definition, or a similarly broad definition, is adopted, many trusts not traditionally regarded as discretionary may fall within the regime. Trustees would need to review their deeds for provisions that could make beneficiaries’ interests technically defeasible, including amendment powers, unit redemption mechanisms and powers to create new classes. Where a trust is caught, trustees would need to amend the deed, restructure using the proposed rollover relief or accept the 30% minimum tax. Each option carries material legal, tax and compliance costs. The burden would be particularly significant for property and infrastructure groups with multiple trusts and complex financing arrangements, as amendments may also require lender consent or trigger financing review events. The period between Royal Assent and the proposed 1 July 2028 commencement may not provide sufficient time to complete this process.

Background to the new proposed tax

On 8 July 2026, Treasury released a Consultation Paper proposing a 30% minimum tax on discretionary trusts, to take effect from 1 July 2028. The stated objective of this measure is to address income splitting, the practice of allocating taxable income to lower-rate beneficiaries to reduce the overall tax burden below the top marginal rate. The proposed measure would impose a 30% minimum tax at the trustee level, with beneficiaries (other than company beneficiaries) receiving a non-refundable tax offset for tax paid by the trustee on their behalf.

The central challenge confronting this reform is definitional. The scope of the proposed minimum tax turns on a single question: what is a ‘discretionary trust’? The Consultation Paper approached this by defining the concept in the negative: a trust that is not a ‘fixed trust’. It invited submissions on whether the existing fixed trust rules in Schedule 2F of the ITAA 1936 provide an appropriate basis for that delineation.

While consultation was available for the proposal, a compressed timeframe of barely three weeks created considerable urgency for industry engagement on what are profoundly consequential design questions. 

Now, in anticipation of the change to the minimum tax on discretionary trusts, it’s important for the industry to examine the definitional challenges Treasury’s proposal raised, and consider:

• whether a generally accepted legal definition of ‘discretionary trust’ exists;
   
• the appropriateness of the existing fixed trust rules as the definitional mechanism;
   
• the exclusions that would be required if those rules are adopted;
   
• an alternative definitional framework better directed at the identified mischief; and
   
• the particular challenges that arise in multi-layered property and infrastructure group structures.

Is there a generally accepted legal and/or tax definition of discretionary trust?

The threshold difficulty confronting the proposed approach is that there is no single, fixed normative definition of ‘discretionary trust’ in Australian law. The concept is essentially descriptive, a label applied to a range of trust arrangements that share certain functional characteristics, rather than a term of art.

Courts have recognised that the term ‘discretionary trust’ is more descriptive than precise, meaning any new law needs to make a policy choice about what should be included. The key question is whether the definition targets the specific issue of income splitting, without unintentionally capturing ordinary commercial trust structures.

Existing state-based definitions show how much the meaning can vary depending on the context. For example, the Victorian Land Tax Act 2005 defines a “discretionary trust” in section 3(1) as:

“a trust under which the vesting of the whole or any part of the trust property (a) is required to be determined by a person either in respect of the identity of the beneficiaries or the quantum of interest to be taken, or both; or (b) will occur in the event that a discretion conferred under the trust is not exercised.”

The Queensland Land Tax Act 2010 adopts a markedly simpler formulation in section 24(2): “discretionary trust means a trust over property for which a person has a power of appointment.”

These definitions were developed for specific land tax purposes and reflect the particular policy concerns of their respective regimes. Neither purports to state a universal test. 

The divergence between these two state-based definitions, operating within the same legal system and for analogous policy purposes, underscores the central difficulty. The concept is context-dependent and varies across statutory regimes.

Use of the existing fixed trust test

In practice, income-splitting is traditionally associated with family trusts and individual family groups or related family group entities. Lower-rate beneficiaries may be family members that benefit from lower marginal rates on the basis that their taxable income for the year has not exceeded the threshold for the top marginal rate. This is a narrower and more precisely targeted concern than is sometimes appreciated. It is directed not at the mere existence of trustee discretions, which are ubiquitous in modern trust practice, but specifically at the ability of a trustee to determine the share of trust income to which beneficiaries are entitled in each income year.

This distinction is important, because it raises a threshold unresolved conceptual question: is this reform properly concerned with ‘discretionary trusts’ as that concept is conventionally understood, or with trusts that merely contain discretions? A trust may contain extensive administrative and management discretions, powers to amend the deed, add or remove beneficiaries, create new classes of units, invest in particular asset classes, without those discretions bearing in any way on the allocation of taxable income among beneficiaries. The opposite of a ‘fixed trust’ is not necessarily a ‘discretionary trust’.

Most trusts that would be classified as ‘non-discretionary’ in ordinary parlance, including widely held unit trusts, property trusts, and infrastructure vehicles, contain discretions of various kinds. The critical question is whether those discretions impact the allocation of taxable income. In the typical unit trust, unitholders are proportionally and presently entitled to trust income by reference to their unitholding. No streaming occurs; the trustee cannot direct income to one unitholder rather than another. The discretions that exist are administrative or structural, not allocative.

Impact on property sector

The property sector would be significantly impacted by this proposed change. Allocation of taxable income may arise where a unitholder redeems units, requiring the trustee to dispose of an asset to fund the redemption, and the trustee allocates the CGT consequences of that disposal to the redeeming unitholder. This limited and transactional exercise of a specific power, directed at matching the tax consequences of a disposal with the beneficiary who economically bears them, does not encounter the same issues as income-splitting. However, the trust would still considered a ‘discretionary trust’ for the purposes of a minimum tax directed at income splitting.

Unless the regime includes an appropriate exclusion or narrower definition, property trusts that allocate CGT consequences to redeeming unitholders may be treated as discretionary trusts. They would need to either remove that discretion, potentially requiring gains to be allocated across all unitholders, or accept the 30% minimum tax on all taxable income, not just the redemption-related gain. Removing the discretion would depart from established market practice, may produce unfair outcomes for continuing unitholders and could require deed amendments, unitholder approval and consideration of resettlement risks. Given that these discretions are used to match tax outcomes with the unitholder who economically triggered the gain, applying the minimum tax would be disproportionate and practically unrelated to income splitting. Property groups should review their redemption mechanisms and continue to advocate for these discretions to be expressly excluded.

The ‘vested and indefeasible’ standard

The existing fixed trust rules are contained in section 272-5 of Schedule 2F to the ITAA 1936. Under that provision, a beneficiary has a ‘fixed entitlement’ to a share of trust income or capital only if the beneficiary has a ‘vested and indefeasible interest’ in that share. Section 272-65 then provides that a trust is a ‘fixed trust’ if persons have fixed entitlements to all of the income and capital of the trust.

This is an unusually high standard. The requirement of indefeasibility means that any power or discretion that could theoretically affect the beneficiary’s interest may be sufficient to render the interest defeasible and thereby disqualify the trust from fixed trust status. In practice, unit trusts with proportionally fixed entitlements may fail the test because trustees typically retain powers that are commercially necessary, the power to amend the deed, to add new unit classes, to exercise administrative discretions but which technically render interests ‘defeasible’ within the meaning of the provision.

Consider a private unit trust holding commercial property for four institutional investors, each entitled to 25% of the trust’s income and capital. Although the trust operates entirely on a proportional basis and does not facilitate income splitting, standard deed provisions, such as powers to amend the deed, create new unit classes and determine redemption values, may make the investors’ interests technically ‘defeasible’. The trust could therefore be classified as discretionary despite operating as a fixed, non-discretionary vehicle. This issue is likely to affect many unlisted property trusts, infrastructure funds and private equity feeder structures, where these powers are included for commercial and governance purposes rather than income splitting.

The Commissioner acknowledged the severity of the ‘vested and indefeasible’ test through Practical Compliance Guideline PCG 2016/16, which sets out circumstances in which the Commissioner will accept that certain discretions do not compromise fixed trust status. However, a practical compliance guideline is not law – it provides administrative comfort, not legal certainty. It may be withdrawn or narrowed at any time and so cannot serve as the definitional foundation for a measure imposing a 30% minimum tax.

Parliament has already demonstrated that a functional approach to fixed trust classification is workable. Under section 276-55 of Division 276 of the ITAA 1997, Attribution Managed Investment Trusts (AMITs) are treated by statute as fixed trusts, with members taken to hold vested and indefeasible interests in the AMITs’ income and capital. This is on the basis that member interests are clearly defined, not because the technical Schedule 2F test is satisfied. PCG 2016/16 acknowledges the range of circumstances that can cause interests to be defeasible, underscoring why administrative guidance is no substitute for a purpose-built statutory test. 

There is no principled reason why a minimum tax directed at income splitting should apply a stricter standard than Parliament has seen fit to apply to equivalent commercial investment vehicles.

Risks of adopting the existing fixed trust rules

Adopting the existing fixed trust rules as the definitional mechanism for the proposed minimum tax creates several risks:

• Excessive breadth. The ‘vested and indefeasible’ standard is extremely strict. Many commercial trusts with genuinely fixed entitlements, property and infrastructure unit trusts with hundreds or thousands of unitholders, would fail the test due to the presence of routine administrative discretions that have no bearing on the allocation of taxable income. These trusts do not engage in income splitting and should not be within scope.
   
• Inconsistency with modern trust practice. Modern trust deeds routinely include amendment powers, add-beneficiary powers, and the capacity to create new unit classes. These are commercial necessities, not indicators of income-splitting purpose. A test that treats their existence as disqualifying fails to reflect the reality of contemporary trust administration.
   
• Uncertainty for widely held commercial structures. Unit trusts with hundreds or thousands of unitholders, which are self-evidently not vehicles for income splitting, may nonetheless be classified as ‘discretionary’ if any power in the deed renders interests technically defeasible. The resulting uncertainty would be commercially unacceptable for major investment platforms.
   
• Difficulty of application in multi-layered structures. In property and infrastructure groups with sub-trusts, the fixed trust test would need to be applied at every level of the structure. Failure at any level cascades through the group, potentially subjecting an entire structure to the minimum tax because of a single administrative power in a sub-trust deed. This is disproportionate to the policy objective.

Preparing for the minimum tax

Businesses operating through trusts should assess their exposure before the proposed 1 July 2028 commencement date.

• Review trust deeds. Identify powers that could make beneficiaries’ interests defeasible, including amendment, unit issue, redemption, forfeiture, valuation and distribution powers.
   
• Map group structures. Review every trust in a multi-layered structure and assess whether one trust’s classification could affect the broader group.
   
• Assess remediation options. Consider whether problematic provisions can be amended, having regard to resettlement and duty risks, financing arrangements, approval requirements and trustee duties. Amendments should not be made before the legislation is settled or without specialist advice.
   
• Model the impact. Compare the cost of restructuring with the potential 30% minimum tax and related limitations on beneficiary offsets.

Proposed exclusions based on using the fixed trust definition

If the existing fixed trust rules in Schedule 2F are adopted as the starting point for defining the scope of the minimum tax, a substantial body of exclusions would prevent the measure from capturing structures that fall outside its policy purpose. 

For example, the following categories of trust could be excluded to ensure the changes meet their objectives:

• Unit trusts with proportionally fixed income entitlements. Where unitholders are entitled to trust income proportionally by reference to their unitholding, and the trustee has no power to allocate income non-proportionally, the trust does not engage in income splitting regardless of whether other discretions exist.
   
• Widely held trusts. The Consultation Paper acknowledges this exclusion, but its scope requires careful delineation, particularly as to the threshold for ‘widely held’ status and whether the exclusion applies where the trust is widely held at one level but has concentrated ownership at another.
   
• Wholesale or institutional unit trusts with qualified holders. Trusts whose unitholders are limited to complying superannuation funds, life companies, and trustees of other widely held trusts do not present income-splitting risks. These investors are either themselves subject to concessional rates (superannuation funds at 15%) or are taxable entities in their own right.
   
• Wholly-owned sub-trusts within excluded groups. Where a head trust qualifies for exclusion, its wholly-owned sub-trusts should equally be excluded. Income-splitting cannot occur at the sub-trust level if the ultimate beneficial ownership is through an excluded entity.
   
• Stapled structures and infrastructure trusts. Where income entitlements are fixed by reference to stapled securities, the trust does not engage in income splitting. The income allocation follows from the security structure, not from any exercise of trustee discretion.
   
• Trusts required by regulation or licence to operate with fixed entitlements. Certain trusts, including registered managed investment schemes, are required by their Australian Financial Services Licence (AFSL) conditions, constitution, or regulatory obligations to maintain proportional entitlements. These should be excluded on the basis of regulatory compulsion.
   
• Fixed trusts (as properly defined). If a new definition of ‘fixed trust’ is developed, one that focuses on the practical ability to allocate income non-proportionally rather than the theoretical defeasibility of interests, trusts satisfying that definition should be excluded.
   
• Insolvency remote special purpose vehicles. Trusts are frequently used in securitisation or structured finance transactions, including for the purpose of issuing securities backed by a pool of receivables. Such trusts are already identified as being within a special category for the purposes of various tax rules, including thin capitalisation. Such trusts are not the subject of income-splitting concerns and should be excluded.
   
• Employee share trusts. Trusts established for employee incentive purposes, including holding employee share scheme interests that are allocated or to be allocated to employees of a corporate group, may include terms and discretions to facilitate the proper functioning of a share scheme, acquisition of shares, and allocation of shares to individual employees. Such trusts are not traditionally the subject of income-splitting concerns and should be excluded.

The scope of the required exclusions is telling. Where the underlying test is so broad that it captures the majority of commercial trust structures, none of which engage in the targeted issue and requires an extensive and complex web of carve-outs to produce a tolerably targeted result, the question must be asked whether the definitional approach itself is misconceived. A better-targeted primary definition would reduce the need for exclusions and produce a more coherent and administrable regime.

An alternative, new definition of discretionary trust

The difficulties identified above suggest that an alternative definitional approach is required, one that is purpose-built for the minimum tax and directed squarely at income-splitting, rather than one that repurposes a test designed for a different statutory context with different policy objectives. 

An adequate definition should focus on substance rather than form. It should ask whether the trustee has the ability to allocate taxable income non-proportionally to a defined group of beneficiaries, the functional hallmark of income splitting.

Core substantive criteria

The definition should capture any trust in which, as a result of the trust deed or any other document, agreement, understanding or unilateral decision, the trustee is able, at any time, to direct, cause or facilitate the allocation of taxable income and/or tax attributes other than on a proportional basis to a defined and ongoing group of persons.

The concept of a ‘defined and ongoing group of persons’ should denote a group of identifiable persons whose interest in taxable income and/or tax attributes has been fixed and is proportional. Their interest in taxable income could not be changed in any way by the trustee, the appointor or any other person involved in the establishment or management of the trust. Critically, the composition of the group should be capable of change only through death, divorce or other operation of law, not through the exercise of trustee or appointor discretion.

The organising principle is that the definition should be directed at the ability to allocate income non-proportionally. Where a trust's constitution fixes the proportional entitlements of a defined group and the trustee lacks the power to alter those proportions, the trust should fall outside the scope of the minimum tax, regardless of what other administrative or protective discretions the deed may contain.

What any adequate exclusion should address

Any exclusion framework should recognise that widely held unit trusts, in which unitholders have subscribed for units and the trust constitution allocates income, taxable income, tax attributes and capital proportionally, without any discretion on the part of the trustee to alter that proportional allocation, do not present the income-splitting issue at which the minimum tax is directed. Such trusts should be excluded at the primary definitional level rather than treated as carve-outs requiring affirmative proof of eligibility.

The exclusion framework should also address the two most common scenarios in which commercial trusts exercise discretions without any income-splitting purpose or effect:

1. Deed amendment powers. A power to amend the trust deed, whether exercised unilaterally by the trustee or with unitholder approval, should not, of itself, bring a trust within scope. The mere existence of an amendment power does not enable income splitting; what matters is whether and how that power is exercised. The consequences of any particular amendment can be assessed at the time of exercise, rather than the bare existence of the power tainting the trust's classification from inception.
    
2. CGT allocation on redemption. The exercise of a discretion to allocate the CGT consequences of an asset disposal undertaken to facilitate the redemption of a unitholder's units should similarly fall outside the scope of the definition. This is a transactional discretion that matches tax consequences with economic outcomes, it does not involve the allocation of the trust's general income stream on a non-proportional basis.

A definition built around these substantive criteria would be more fit-for-purpose than the existing fixed trust rules.

Practical implications for those using complex structures

The Consultation Paper's proposed framework raises particular challenges for property and infrastructure groups that employ multi-layered trust structures, a common and commercially necessary arrangement in the Australian market. For groups with a mixture of resident, non-resident, and superannuation fund unitholders together with wholly-owned sub-trusts, the proposed minimum tax creates not merely technical classification questions but immediate commercial and operational risks that require active management. Groups in this position should be assessing the following interlocking difficulties now, with a view to quantifying their financial exposure, identifying structural vulnerabilities, and determining whether advocacy, restructuring, or a combination of both is the appropriate response.

Double tax risk

The most immediate concern is the risk of double taxation. If a head trust qualifies for an exclusion from the minimum tax (for example, because it is widely held or because its unitholders have proportionally fixed entitlements), but its wholly-owned sub-trust does not independently satisfy the same test, the sub-trust would be subject to the 30% minimum tax. Income would be taxed at 30% at the sub-trust level before flowing up to the head trust and being distributed to unitholders. Without a full and effective offset mechanism that flows through the trust chain, this produces double taxation that has no nexus to income splitting.

Interaction with withholding tax exclusion

The Consultation Paper provides that distributions to foreign residents comprising dividends, interest, or royalties subject to withholding tax are excluded from the minimum tax. This exclusion recognises that such income is already subject to final withholding tax and should not bear an additional 30% impost. However, the exclusion is limited to those specific income types. Other income types commonly distributed to non-residents by property and infrastructure trusts, rental income, capital gains from the disposal of Australian real property, and income from cross-staple arrangements, may not fall within this exclusion.

The consequence is that a sub-trust distributing rental income or real property capital gains to a non-resident unitholder may be subject to the 30% minimum tax on that distribution, notwithstanding that the non-resident is separately assessable on the income under the ordinary provisions and the new foreign resident CGT regime. The interaction between the minimum tax and the expanding foreign resident CGT base requires careful consideration to avoid overlapping imposts.

Superannuation fund unitholders

Complying superannuation funds are excluded from the minimum tax, a necessary recognition that their concessional 15% rate reflects a deliberate policy choice, not income splitting. However, where a sub-trust has both superannuation fund unitholders and non-superannuation entity unitholders, the interaction between the 30% minimum tax and the 15% concessional superannuation rate creates considerable complexity.

If the sub-trust is classified as a discretionary trust (because it fails the fixed trust test), the 30% minimum tax applies to all distributions other than those to excluded beneficiaries. The superannuation fund is excluded, but the non-superannuation unitholders bear 30% minimum tax on their share. This may be an appropriate outcome in a genuine income-splitting context, but where the sub-trust has proportionally fixed entitlements and the allocation is non-discretionary, it would be an unintended consequence. 

Deed amendment and restructuring costs

If the fixed trust rules in their current form are adopted, strict application would require many groups to amend trust deeds to remove routine administrative discretions that currently render interests technically defeasible. Deed amendment in the trust context requires legal advice on the scope and effect of the amendment, consideration of potential resettlement risk (which, if triggered, would constitute a CGT event and may give rise to stamp duty consequences), and compliance with any procedural requirements in the deed or under general law. This would introduce substantial costs. 

For large groups with multiple trusts, there is an aggregate cost of deed amendment and the associated risk assessment. This cost is particularly concerning where the amendments remove discretions that have no bearing on income allocation and whose presence reflects standard commercial drafting practice rather than income-splitting design.

A ‘look-through’ or group-based approach

For property and infrastructure groups, the application of the fixed trust test at each level of the structure independently would present challenges. A ‘look-through’ or group-based approach, where exclusion at the head trust level cascades to wholly-owned sub-trusts within the group, would be more practical.

This is because in a structure where a widely held head trust with proportionally fixed entitlements wholly owns a sub-trust, income splitting does not and cannot arise. The economic substance is one of proportional fixed entitlements flowing through a group structure. The sub-trust’s income is attributable to the head trust’s unitholders proportionally; no trustee at any level has the capacity to allocate income non-proportionally to achieve a tax advantage.

A group-based approach would provide that where:

1. head trust satisfies the exclusion criteria (whether as a widely held trust, a trust with proportionally fixed entitlements, or otherwise);
    
2. the head trust wholly owns (directly or indirectly) one or more sub-trusts; and
    
3. the income of those sub-trusts flows to the head trust and is distributed proportionally to its unitholders,
    
4. the sub-trusts are taken to satisfy the exclusion criteria by reason of their membership of the group, without the need for independent testing at each level.

This approach is consistent with the stated policy objective of the Consultation Paper. It would produce a result that is targeted at income splitting, commercially workable, and proportionate in its compliance burden. It avoids double taxation, reduces deed amendment and restructuring costs, and provides certainty for groups whose structures do not engage in income-splitting.

Next steps

The definition at the heart of the Consultation Paper is not merely technical. It will determine whether the measure achieves its stated anti-income-splitting objective or whether it inadvertently captures broad swathes of commercial trust structures, unit trusts, property vehicles, infrastructure groups, stapled entities, structured finance vehicles, and employee incentive vehicles that do not engage in the behaviour the measure seeks to prevent. 

Getting the definition wrong could impose significant costs on the Australian funds management industry, deter investment in Australian real property and infrastructure, and create ongoing uncertainty for hundreds of thousands of commercial trusts.

For now, businesses with trusts that fall under the existing proposed definition should consider:

• Auditing existing trust structures against the proposed definition;
   
• Identifying trusts with administrative discretions that could cause them to be caught; and 
   
• Seeking advice on whether restructuring may be needed before the 1 July 2028 start date.