Australia Has Repriced Its Largest Asset Class.

The findings, before the argument.

On 7 September 2026 the Australian Financial Review Property Summit put a number on the money leaving Australian residential property, and James Thomson's Chanticleer column the same day carried it under the headline “How ‘seismic’ housing crash could unleash $3.6trn shock across property”. Charter Hall's David Harrison told the summit that about A$3.6 trillion is invested in residential property, most of it by households, and that the May tax changes will push it towards commercial property; Morgan Stanley's Tim Church said prices could fall as much as 15 per cent from peak to trough, the largest fall on record. That commentary is about where the money goes next. This paper is about whether it can get there.

• The A$3.6 trillion being quoted is the value of investment property held. It is not money about to move.
• Australia's May budget barely moved house prices. It changed which legal structure it makes sense to own an investment property in.
• The average Australian residential investor was already losing money. Losing the salary offset is worth a 90 to 155 basis point rise in mortgage rates.
• A superannuation fund is now the most tax-favoured way for an Australian to hold a rental property, and the same Act bans self-managed funds from borrowing to buy one.
• The wealth tests that decide who may be sold wholesale investments in Australia were set in 2002 and never indexed. House prices have widened them instead.
• Self-managed superannuation funds, Australia's self-directed pension vehicles, hold A$1.06 trillion, and the regulator and the complaints authority disagree on whether their trustees count as wholesale investors.
• Australian private credit, the obvious home for this money, has just frozen redemptions after a A$3.4 billion developer collapse.
• Australia's institutional pension funds already invest half their money offshore. Displaced household capital has no domestic product built to receive it.

The A$3.6 trillion being quoted this week is the value of investment property held, not money about to move. Between a balance-sheet total and any Australian destination sit four further tests: whether the property can be sold without a worse outcome than holding, whether it actually turns over, whether the proceeds sit in a structure that can buy something else, and whether that structure is legally allowed to be offered the product. Treating the total as if it had passed all four turns a number that commits nothing into an imminent wave. The last test is the one nobody is examining.

Australia's federal budget of 12 May 2026 did not move house prices much. The Australian Treasury's own modelling says around 2 per cent over a couple of years; the Commonwealth Bank of Australia says just under 3. What it changed is which legal structure it makes sense to hold residential property in. Rental losses can no longer be set against salary. The individual capital gains discount is gone, replaced by indexation and a 30 per cent minimum rate. Superannuation funds — Australia's pension vehicles — keep their discount.

The cohort it landed on was already losing money. On the Australian Taxation Office's latest figures, 2.33 million individuals declared net rent averaging −A$1,148, and 54.2 per cent of rental investors were negatively geared — up from 49.4 per cent a year earlier. Removing the salary offset is worth, in cash-flow terms, a 90 to 155 basis point rise in investor mortgage rates, permanently, for anyone holding in a personal name.

Read the Act as a whole and it contradicts itself. The structure it now favours for holding residential property — a superannuation fund — is the structure it forbids from borrowing to buy one, from 10 August 2026, by an Australian Greens amendment that was not in the budget. The other structure a displaced investor would reach for, the discretionary family trust, is next: a 30 per cent minimum tax from 1 July 2028, with roughly 840,000 of them in the frame.

Who may be offered a wholesale investment is decided by Australian wealth tests set in March 2002 and never indexed: a A$500,000 parcel, A$2.5 million of net assets, A$250,000 of income. The Australian Securities and Investments Commission (ASIC) has asked for them to be raised to roughly A$922,000, A$4.61 million and A$461,000. Rising house prices have done the widening the Australian Parliament never legislated — 1.9 per cent of Australian adults met the tests in 2002, 43.6 per cent are forecast to by 2041 — and in February 2025 a parliamentary committee declined to lift the numbers.

The largest pool of household capital that could be redeployed, 672,805 self-managed superannuation funds holding about A$1.06 trillion, sits inside a contradiction that has been unresolved for more than a decade. ASIC, the corporate regulator, says it will not act against a provider that treats a fund's trustee as a wholesale investor on the ordinary wealth tests. The Australian Financial Complaints Authority says the law makes that trustee a retail client unless the fund holds A$10 million, and will hear complaints on that basis. For a manager, that is the difference between a distribution channel and a compliance exposure.

The obvious domestic home for this money — Australian private credit, around A$200 billion of it — has just frozen redemptions in the same season it was asked to receive it. The Sydney developer Universal Property Group went into administration owing roughly A$3.4 billion, some 40 funds are exposed, and three fund managers — Centuria, MA Financial and CVS Lane — have limited withdrawals. ASIC found four funds in 28 that publish the rates they charge borrowers.

The institutions left first: half of Australia's professionally managed retirement savings is already invested offshore. Household capital has now been pushed out of the asset that anchored it at home. Both are looking for a destination at the same time, one has already chosen, and no policy currently in force gives the other a reason to behave differently. The argument that follows takes the four tests in order and ends with the sequence of work for anyone proposing to receive this capital.

Displacement is not deployment.

Two accounts of the Australian housing correction are circulating this week, and both are incomplete. The first holds that Australia's May budget has broken the residential investment market and that the damage is a price event. The second — put most forcefully by Charter Hall's David Harrison at the Australian Financial Review Property Summit on 7 September, and carried by Chanticleer as a “seismic” $3.6 trillion shift — holds that the capital released will rotate into commercial property, and that the rotation is a matter of time. Neither survives contact with the plumbing.

What Australia's Treasury Laws Amendment (Tax Reform No. 1) Act 2026 did was not principally to change the price of Australian residential property. It changed the legal form in which holding it makes sense. That is a different kind of event, and it produces a different kind of problem. A price event resolves itself: assets reprice, buyers return, the market clears. A change of form does not resolve itself, because the capital has to move through a legal structure to get anywhere, and the structures available to receive it in Australia are either newly disadvantaged, prohibited from the obvious transaction, or gated by financial thresholds that have not moved since March 2002.

The number being quoted this week is $3.6 trillion. It is being used as though it describes capital about to move — Harrison's own framing was that “if even 10 per cent of that capital”, or $360 billion, starts to shift, commercial property has “a huge demand driver”. It does not. It describes a stock, held in a particular legal form, most of which will not turn over in any given year. Between that stock and any Australian destination sit four further states, and the last of them — legal eligibility to subscribe — is where the argument is, and where nobody is looking.

Every figure below was checked at its publishing source. All statements of position are as at 8 September 2026. All dollar figures are Australian dollars.

What Australia's 12 May budget actually did

The measures were announced in the Australian federal budget for 2026–27 on 12 May 2026 and are law. The Treasury Laws Amendment (Tax Reform No. 1) Act 2026, No. 49 of 2026, received Royal Assent on 26 June 2026. The commentary has run ahead of the statute in two places, and both matter.

Negative gearing — the deduction of rental losses against other income, including salary — was quarantined, not abolished. New section 26-155 of the Income Tax Assessment Act 1997 provides that where deductions relating to holding residential dwellings exceed assessable income from residential dwellings, the excess "is not deductible for that income year" but becomes "a quarantined amount" that carries forward indefinitely. It can still be applied against residential rental income and against residential capital gains. It cannot be applied against wages, and it cannot be applied against commercial property income. Losses are not destroyed. They are severed from salary.

The capital gains tax discount was replaced, not removed. For individuals, trusts and partnerships, the 50 per cent discount gives way to indexation of the cost base for CPI, together with a 30 per cent minimum tax rate on the real gain, for gains accruing on or after 1 July 2027. Companies never had the discount. Superannuation funds keep theirs — the one-third discount is untouched.

The dates are widely misreported. 12 May 2026, 7.30pm, is an acquisition cut-off: an interest last acquired before that moment is grandfathered and remains negatively gearable until sold. The quarantining rule itself applies from the 2027–28 income year. A property contracted on 1 June 2026 is negatively gearable in the ordinary way until 30 June 2027, and quarantined thereafter.

The Australian Treasury's own modelling is modest about the price effect: prices growing "around 2 per cent less over a couple years" than they otherwise would, roughly 75,000 additional owner-occupiers over a decade, and a rent effect of "less than $2 per week" on the median. The Commonwealth Bank of Australia, publishing the day after the budget, put the price effect at just under 3 per cent below baseline and made the more useful observation: removing the salary offset is equivalent in cash-flow terms to an increase in investor mortgage rates of roughly 90 to 155 basis points.

That equivalence is the point. It is not a valuation shock. It is a permanent increase in the cost of holding a residential investment property in a personal name, imposed on a cohort whose average position was already a loss. The Australian Taxation Office's 2023–24 statistics record 2,332,653 individuals declaring net rent at an average of −$1,148 and a median of −$934. The average Australian residential property investor was, on the most recent published data, losing money before the change.

The one structure the tax code now favours is the one forbidden to borrow.

Read the two schedules together and an asymmetry appears that has attracted almost no comment.

Schedule 2 excludes complying superannuation entities, including self-managed superannuation funds (SMSFs, Australia's self-directed pension funds), from the negative gearing quarantine. Schedule 1 leaves the superannuation CGT discount intact while removing the individual one. On the face of the Act, the Commonwealth has just made the superannuation fund the most tax-advantaged vehicle in which an Australian can hold a residential investment property, relative to every alternative available to a household.

Schedule 5 of the same Act then prohibits self-managed superannuation funds from using limited recourse borrowing arrangements to acquire residential property, with effect from 10 August 2026. That schedule was not in the budget. It was accepted as an Australian Greens amendment in exchange for passage.

The position in one line — The structure the tax code now favours for holding residential property is the structure forbidden from borrowing to buy it. A household that responds rationally to Schedule 1 and Schedule 2 runs directly into Schedule 5.

The effect is not that capital moves into superannuation and buys houses. The effect is that the tax system has closed the personal-name route into leveraged residential property and simultaneously closed the geared superannuation route, leaving the displaced household investor with an unleveraged choice between residential property held on worse terms than before, and something else entirely.

And discretionary trusts, the other structure a displaced household investor would reach for, are next. The same budget announced a 30 per cent minimum tax on discretionary trust taxable income from 1 July 2028, costed at $4.5 billion over five years, with a three-year restructure rollover from 1 July 2027. Exposure draft legislation was out for consultation from 3 to 18 September 2026. There are roughly 1.06 million trusts lodging returns in Australia and, on Treasury's own figures, around 840,000 of them are discretionary.

The institutions left first.

This is the second half of an argument set out in Where the Money Goes, published on 6 September 2026. The finding there was that roughly half of Australia's institutionally managed superannuation pool is already deployed outside the country, that the NAB Super Insights survey recorded an international allocation of 50.9 per cent for 2025 — the first print above half — and that funds expect 50 to 60 per cent of each incremental dollar to keep going abroad. The Future Fund holds approximately 77 per cent of its assets offshore on a look-through basis.

That was not sentiment. It was section 52 of the Superannuation Industry (Supervision) Act 1993 applied to a domestic opportunity set that is around two per cent of world equity indices and concentrated in banks and resources. Trustees have no statutory permission to overweight Australian assets that do not clear the same risk-adjusted hurdle.

So the sequence is now complete. The institutional capital left because the domestic opportunity set did not reward it. The household capital has now been pushed out of the asset that anchored it at home. Both are looking for a destination at the same time, and one of them has already chosen.

Five states, one number

The displacement is persistently discussed as a flow because the language used to describe household capital does not distinguish between five materially different states. A stock is not realisable. Realisable is not mobile. Mobile is not redeployable. Redeployable is not eligible. Each compression is defensible in isolation. Together they turn a balance-sheet total into an imminent wave.

The $3.6 trillion attributed this week to Charter Hall's David Harrison, speaking at the Australian Financial Review Property Summit on 7 September 2026, is a figure at the first rung. It describes residential investment property held, and Harrison's own condition — “if even 10 per cent of that capital” moves — is the point. The Australian Bureau of Statistics puts the entire dwelling stock at $12,772.6 billion for the March quarter 2026, and the Reserve Bank of Australia's May 2026 Bulletin analysis puts investment properties at around 20 per cent of the dwelling stock by number. The Bank is explicit that this is a count, not a share of value, and that the series covers individuals only, excluding partnerships and trusts. The two figures cannot be reconciled without an assumption about the relative price of investor-held dwellings that the Bank's own data does not support. That is not a criticism of the figure. It is a caution about which rung it sits on.

Note on timing: the ABS Total Value of Dwellings release for the June quarter 2026 is scheduled for 8 September 2026 and will be the first stock valuation to capture the decline Cotality has recorded since March. The March quarter figures above are stated with their reference period for that reason.

The rung nobody has priced

A household that sells an investment property and wants an income-producing replacement in Australian private markets — a mortgage fund, a credit fund, a wholesale scheme — has to be legally permitted to be offered it. That permission is governed by financial thresholds set in regulations that commenced in March 2002 and have never been indexed.

Three consequences follow, and they run in opposite directions.

The gate has widened by accident, and the regulator wants it narrowed

The Australian Securities and Investments Commission (ASIC) told a Parliamentary Joint Committee in May 2024 that the thresholds "have not been increased since their introduction into the Corporations Act in the early 2000s", and that adjusting them for CPI from 2001 to 2024 would lift the product value test from $500,000 to approximately $922,000, the net asset threshold from $2.5 million to approximately $4.61 million, and the gross income threshold from $250,000 to approximately $461,000. It recommended exactly that, plus a mechanism for periodic increases.

The scale of the drift is in research by Ben Phillips at the Australian National University's Centre for Social Research and Methods, cited in the same ASIC submission: the share of Australian adults meeting the individual wealth tests was 1.9 per cent in 2002, was predicted to reach 16.2 per cent by 2021, and is forecast at 43.6 per cent by 2041. Nobody legislated that. House prices did it.

The Parliamentary Joint Committee reported in February 2025 with two recommendations: that government consider a mechanism for periodic review, and that the subjective elements of the sophisticated investor test be replaced with objective criteria. It did not recommend raising the financial thresholds. As at today they stand unchanged.

The largest single pool of redeployable household capital is the one whose status is contested

Section 761G(6) provides that where a financial service relates to a superannuation product, the trustee of a superannuation fund is a retail client unless the fund has net assets of at least $10 million, and that the section 761G(7) tests do not apply in that case. ASIC withdrew its 2004 guidance in August 2014 and adopted a no-action position: it "will not take action" where a provider applies the general test to a trustee subscribing for financial products on behalf of an existing fund. In the same release it said, in terms, that this "will not affect any private rights of action that may be available to third parties" and that providers "need to make their own commercial decisions after considering the legal risks."

In June 2025 the Australian Financial Complaints Authority set out the other half. Its Lead Ombudsman and a Senior Ombudsman wrote that "under the law, if an advisor provides advice to a trustee in relation to an SMSF, it must be treated as a retail client unless the SMSF has $10 million or more in assets", that ASIC's position "does not limit consumer action", and that an AFCA panel has already found a wholesale classification incorrect. Investor sophistication, they wrote, may reduce compensation; it does not alter classification.

The regulator will not prosecute. The complaints body will take jurisdiction. Both positions are published, both are current, and they point in opposite directions.

The arithmetic makes this material rather than technical. There are 672,805 self-managed superannuation funds holding approximately $1.06 trillion at the March 2026 quarter, which is an average of about $1.58 million a fund. On the $10 million reading, almost the entire sector is retail whatever the trustees' experience. On the $2.5 million reading, a substantial minority qualifies. The difference between those two readings is the difference between a distribution channel and a compliance exposure, and it has been unresolved for more than a decade.

Family trusts are the clean route, and they are the next target

There is no superannuation-style carve-out for trustees of family or discretionary trusts. They fall under the ordinary tests, and section 708(8)(d) expressly extends the Chapter 6D exemption to a company or trust controlled by a qualifying person, with subsections 708(9B) and (9C) allowing the controlled entity's own net assets and income to be counted. A household that owns a home and an investment property clears $2.5 million of net assets routinely. The very property equity now being repriced is what makes the holder eligible.

That is the cleanest available path from a displaced residential holding into a domestic income asset. It is also the structure on which the same budget announced a 30 per cent minimum tax from 1 July 2028.

What else could explain this

An argument built on one budget and one quarter has to deal with the alternatives. Three compete with the account set out here, and each would be preferred to it if the evidence supported it.

This is the rate cycle, not the tax change

The most serious objection, and partly correct. The Reserve Bank of Australia raised the cash rate three times in 2026 — 4 February, 18 March and 6 May — taking it from 3.60 to 4.35 per cent, and held in June and August. The ten-year Australian government bond yield reached a fifteen-year high in early September. Australian house prices would be falling on that alone.

But rates do not explain the composition of what is happening. A rate cycle depresses prices across property types; it does not sever residential losses from salary income while leaving commercial deductions intact, and it does not remove the individual CGT discount while preserving the superannuation one. Those are statutory choices, and they change which legal form a holding makes sense in — which is the whole of the argument here. The Reserve Bank itself, in its own words on 11 August 2026, records that "momentum in the housing market has shifted, with housing prices falling in some capital cities and new housing loans declining noticeably." It does not attribute that to itself alone.

The money will go to listed equities and ETFs, where no plumbing is needed

A fair reading, and the likeliest single destination. Listed markets require no certificate, no threshold and no structure. Australian self-managed funds already hold 26 per cent of their assets in listed shares and 16 per cent in cash and term deposits.

Two things argue against it as a complete account. The first is that the displaced holder is not seeking growth; that is precisely the exposure the tax change has made unattractive. It is seeking income, and a fifteen-year high in bond yields sets a hurdle that listed equity income does not clear on a risk-adjusted basis for a holder in pension phase. The second is that the objection concedes the point: if the plumbing is the reason capital defaults to listed markets, the plumbing is the binding constraint on everything else, which is the argument.

Nothing will move, because the grandfathering is generous and turnover is low

Also partly correct, and the reason this is a multi-year argument rather than a 2026 one. An interest acquired before 7.30pm on 12 May 2026 keeps its treatment until it is sold, which is an incentive to hold. Cotality, the property data provider, records sales volumes 15.5 per cent below a year earlier.

But grandfathering attaches to the interest, not to the investor. It does not survive a sale, a death, a divorce or a restructure, and it does nothing for the next acquisition. The stock does not have to move for the flow to change: it is sufficient that the roughly 230,000 individuals a year who, on the Australian Treasury's figures, acquired a negatively geared property no longer do so. The displacement is at the margin, and the margin is where allocation decisions are made.

What would prove this wrong

If Australian residential investor loan approvals return to their pre-May 2026 trend during 2027 while the quarantining rule is in force, the argument that the change is structural rather than cyclical is wrong. If the government legislates an increase in the wholesale thresholds and the flow of household capital into domestic private markets falls rather than rises, the eligibility argument is wrong. And if ASIC's no-action position is codified, or a court resolves section 761G(6) in favour of the general test, the largest single obstacle identified here disappears. We will publish either way.

Nothing domestic is built to receive it.

Assume the capital clears every rung. Where does it go?

The rotation thesis advanced at the property summit is commercial real estate, and the supply argument behind it is real: Scentre reports around 100 vacancies across the entirety of Westfield's 42 Australian malls; Cushman & Wakefield puts roughly 520 hectares of Melbourne and Sydney industrial land as removed from the industrial market; Charter Hall's Chifley Tower will be one of only two new office developments built in the Sydney CBD in the next five years. But commercial property equity is not an income substitute for a displaced residential holder at household scale, and the listed route into it has been punished: the S&P/ASX 200 A-REIT index returned −15.50 per cent over the twelve months to 31 August 2026 against +4.40 per cent for the S&P/ASX 200.

The natural destination is domestic private credit, and it is having the worst month of its short life.

ASIC's surveillance of 28 private credit funds, published on 5 November 2025, found that only four published information about the interest rates charged to borrowers; that fewer than half had detailed written credit, impairment or default management policies; that most had no effective separation between the investment committee approving loans and those monitoring them; and that of the wholesale funds, only two performed stress testing as part of liquidity risk management. On 18 June 2026 ASIC said the sector "is facing its first real test" and that market participants "should not wait for formal defaults before reassessing asset values and related risks."

Then it arrived. Administrators from Teneo were appointed to the Sydney developer Universal Property Group, the Bathla group, on 25 August 2026, reporting approximately $3.4 billion owed to known creditors at the first creditors' meeting on 4 September. Around 40 private credit funds are reported to have exposure, ranging from $1.5 million to $340 million. Centuria paused redemptions and applications on two credit funds; MA Financial capped redemptions at up to 1 per cent of funds under management a month; CVS Lane temporarily suspended applications and redemptions across two funds with $2.1 billion under management and exposure across nine loans. On 4 September 2026 the ASIC Chair told the Parliamentary Joint Committee that "there have been several troubling developments in the private credit sector, most notably with the recent collapse of Bathla", and that "currently there is a lack of information and insight into wholesale private credit funds."

This is the sector being asked to receive displaced household savings, in the same season in which it gated them.

And the record on who was actually holding the risk is now established. ASIC stated in February 2026 that around 11,000 Australians invested approximately $1.1 billion in the Shield and First Guardian funds, both of which collapsed. In June 2025 it reported that Australian Fiduciaries Limited had raised approximately $160 million from around 600 retail investors "predominantly via SMSFs". ASIC's own capital markets report of November 2025 put the concern plainly: private credit funds "are increasingly targeting retail and less sophisticated wholesale investors, including individuals", and "while these investors may meet the specified asset and income thresholds, many are unable to fully assess the risks."

The collision — The Commonwealth has displaced household capital from its largest asset class. The regulator has asked for the gate into the alternative to be narrowed. The alternative has just gated itself. All three are true simultaneously, and none of the three was designed with reference to the other two.

What would actually redirect it

If the constraint is form and eligibility rather than appetite, the order of work for anyone proposing to receive this capital is not the order most Australian managers are following.

The order matters because stage five has become easy and stage one has become hard. There is more displaced Australian household capital looking for domestic income than there are domestic income products it can lawfully and comfortably hold. A manager arriving at that market with a settled eligibility perimeter and disclosed risk is not competing for scarce money. It is a scarce destination competing for capital that has nowhere to go, and it will be priced accordingly.

The alternative is the one already visible in the institutional data. Half of Australia's professionally managed retirement savings is invested outside the country because the domestic opportunity set did not reward it. There is no law of nature that says displaced household capital will behave differently, and no policy currently in force that gives it a reason to.

Method and sources

Every figure was measured, published or reported between 1 January 2025 and 8 September 2026, and each is stated with its reference period. Figures from earlier periods were excluded except where they date a statutory event or a regulatory position that remains current. Every figure was checked against the publishing source rather than secondary coverage of it. Where a figure is derived rather than published it is marked DERIVED and the arithmetic is shown. Where a figure could not be traced to its source it was removed rather than softened.

Two figures are stated with their limits rather than reconciled. The $3.6 trillion attributed to Charter Hall describes residential investment property held and cannot be reconciled with the ABS dwelling stock and the RBA’s count-based investor share without an assumption neither publisher supports. And the ABS Finance and Wealth series puts household land and dwellings at $12,983.0bn where Total Value of Dwellings puts the household-owned dwelling stock at $12,266.8bn, both at the March quarter 2026. They are different national accounts concepts and are not netted here.

No confidential information and no client material was used. This paper describes markets and legislation. It describes no financial product, and no product of any kind is offered or referred to in it.