Twelve years late, but welcome: the SMSF residential borrowing ban

Since Monday 10 August, a self-managed superannuation fund has been unable to enter a new limited recourse borrowing arrangement to acquire residential property. The Treasury Laws Amendment (Tax Reform No. 1) Act 2026 received Royal Assent on 26 June, and the standard 45-day commencement window did the rest. The regime is now more than two weeks old, and the early questions arriving in practitioners' inboxes suggest the profession's work on it is only beginning.

For a profession that has spent the better part of two decades watching leveraged residential property get sold into superannuation by parties with no advice obligation and no ongoing relationship with the client, this is a good outcome. It is also an outcome that a major inquiry recommended in 2014 and that the Council of Financial Regulators supported in 2019. The question is not whether the change is justified. It is why it took a Senate negotiation on an unrelated tax bill to finally deliver it.

What actually changed

The reform is narrower than the headline suggests, and advisers who read only the headline will give wrong advice.

The amendment, moved by Senator Nick McKim for the Greens on 23 June as the price of Senate passage for the government's broader CGT and negative gearing package, inserts a single new paragraph into the LRBA conditions in section 67A of the Superannuation Industry (Supervision) Act 1993: where the asset acquired is real property, that asset must be business real property within the meaning of section 66. The ATO has been explicit that LRBAs themselves are not banned, and that the restriction confines real property acquisitions to business real property regardless of whether the lender is a bank, a non-bank or a related party.

That framing matters. As Boyce and others have pointed out, the operative test is business real property, not a residential-versus-commercial split, which means some residential property remains eligible and some commercial property does not. Business real property, defined in subsection 66(5) and explained in SMSFR 2009/1, generally requires an eligible interest in land used wholly and exclusively in one or more businesses. A serviced apartment, a mixed-use shopfront with a flat above it, vacant land held for development, a lifestyle block, or a farm with a homestead all now require careful analysis before contracts are exchanged, and the treatment of off-the-plan commercial premises remains unclear pending further guidance. Advisers who have historically treated "commercial equals fine" as a working rule will find that rule has become a liability.

The transitional protection is generous and should calm most existing clients. Arrangements entered into before commencement are grandfathered with no forced sale and no restructure. Refinancing of pre-commencement borrowings remains available. Critically, a binding contract exchanged before 10 August is protected even where the LRBA is entered into and settlement occurs after that date, including off-the-plan contracts settling a year or more out. The ATO has noted that later variations will generally not disturb that protection, though a contract altered so fundamentally that its original terms no longer exist may be treated as a new arrangement. That caveat is now the live risk for the grandfathered cohort, and it belongs in every settlement file.

Why this was overdue

The case against leverage inside superannuation was made clearly, and rejected, more than a decade ago.

The 2014 Financial System Inquiry chaired by David Murray recommended restoring the general prohibition on direct borrowing by superannuation funds by removing the section 67A exception on a prospective basis. The Inquiry's reasoning was twofold: prevent an unnecessary build-up of risk in the superannuation and broader financial systems, and reinforce superannuation as a retirement income vehicle rather than a general wealth management structure.

The then government declined that recommendation in 2015 on the basis that the data was insufficient, and instead commissioned the Council of Financial Regulators and the ATO to monitor leverage and report back in three years. That report landed in February 2019. It found that 8.9 per cent of SMSFs held an LRBA, covering 5.2 per cent of total SMSF assets, or more than $38 billion, and that both the number of funds and the value of leveraged assets were still growing. Nothing happened.

What did happen was that the lenders voted first. NAB withdrew from LRBA property lending in July 2018, restricting SMSF lending to shares and managed funds. AMP and CBA followed within months. The major banks exited a decade ago on credit and reputational grounds, and the market migrated to specialist non-banks. That migration is itself a reason to be comfortable with the ban: an asset class that the country's largest and best-resourced credit assessors declined to fund inside super was never a mainstream retirement strategy.

By September 2025, ATO figures had SMSFs holding roughly $75 billion in assets carrying associated debt of about $28.9 billion, against a sector exceeding $1 trillion. Small in aggregate. Not small at all in the funds concerned, where the arrangement routinely represents the overwhelming majority of the member's retirement savings, held in a single, illiquid, undiversified asset with a geared balance sheet and a pension phase approaching.

That concentration is the real objection, and it is an advice objection before it is a prudential one. A two-member fund with a $900,000 property, a $600,000 loan and $80,000 in cash cannot pay a death benefit, cannot fund a minimum pension in a vacancy, cannot rebalance, and cannot exit without a disposal that is entirely dependent on market conditions at the time the trustees least control. The section 67A exception was drafted for instalment warrants over listed securities in 2007. It became a residential property distribution channel. That was never the design.

The objections worth taking seriously

Two criticisms deserve a hearing, and neither is disposed of by the merits of the policy.

The first is process. The SMSF Association has described the change as a significant reshaping of the SMSF investment landscape delivered through a late-stage amendment without consultation or an evidence-based review process. That criticism is correct on the facts. The ban was the price the Greens extracted for supporting the government's CGT discount and negative gearing package, not a considered piece of superannuation policy. Good outcomes reached badly still leave problems behind, and the business real property drafting is exactly the kind of definitional ambiguity that consultation exists to catch. The test now doing the work was designed for related-party acquisition and leasing rules, not for borrowing, and practitioners are relying on SMSFR 2009/1 and section 66 case law to resolve questions the legislature never turned its mind to.

The second is scale and housing supply. The affected market may be materially larger than official data suggests. Preliminary Australian Finance Industry Association figures from specialist non-bank lenders record more than 16,000 new residential SMSF loans written in FY26, with total security of $10.3 billion, roughly four times the volume implied by ATO data. Daniel Butler of DBA Lawyers has argued that restricting SMSF acquisition of new dwellings reduces supply while leaving build-to-rent concessions available to large domestic and offshore institutional developers. Whether or not one accepts the supply argument, the discrepancy in the data is a fair indictment of a reform passed without an evidence base.

Neither objection rescues the underlying strategy. Both should temper any suggestion that this was well-executed policymaking.

What this means for practice now

For clients with existing arrangements, the message is stability. No action is required, refinancing is preserved, and the correct advice for most is to resist unnecessary restructuring. Note separately that from 30 June 2026 outstanding LRBA amounts are included in a member's total superannuation balance, which interacts directly with the Division 296 threshold that applies from 1 July 2026. Clients near $3 million need that modelled now, not at the next review.

For clients who exchanged before 10 August and are yet to settle, the file needs two things: evidence of the pre-commencement exchange date, and discipline around variations. Extending a settlement date is one thing; renegotiating the substance of the contract is another, and the difference is now the difference between a protected arrangement and a prohibited one. Off-the-plan purchasers settling into 2027 should have this conversation before the developer proposes the first amendment, not after.

For clients who missed the deadline, there is no workaround worth its name, and advisers should be wary of the ones now circulating. The test is contract exchange before 10 August, and an arrangement papered to look older than it is invites exactly the ATO attention the grandfathering was designed to avoid.

For everyone else, the strategic conversation shifts. Ungeared direct property inside super remains available for funds with the cash. Business real property acquisitions remain financeable where the section 66 test is genuinely satisfied, which now warrants written analysis on file rather than an assumption, and the early signs of a pivot toward commercial LRBAs make that analysis the next compliance frontier rather than a formality. And for clients whose objective was leveraged residential exposure inside a concessionally taxed environment, the honest answer is that the structure was always a poor fit for the objective, and the legislation has simply removed the option of pretending otherwise.

The Murray Inquiry was right in 2014. The Council of Financial Regulators and the ATO confirmed it in 2019. It should not have taken a crossbench trade, twelve years and one accidental deadline to act on it.