The 2018-2020 SMSF lender withdrawal — and why it still shapes the market.

It's easy to miss that there was a moment, less than a decade ago, when buying property through a self-managed super fund involved a competitive choice between major-bank lenders. In 2016 a Sydney-based SMSF trustee buying an investment property could compare CBA, Westpac, NAB, AMP and Macquarie on similar terms — major-bank pricing, established LRBA products, broad LMI access. By mid-2020 that menu didn't exist. The major banks were gone. The lenders remaining in the residential SMSF segment were a small group of non-bank specialists at materially higher pricing.

The withdrawal happened quickly enough that many borrowers didn't notice until they tried to refinance. It happened quietly enough that none of the involved banks made a single high-profile announcement. And it happened consistently enough across the major-bank tier — within roughly 18 months — that it can't be explained as four independent commercial decisions. Something structural was moving the panel.

The timeline

The sequence isn't disputed in industry reporting; the cluster of withdrawals is a matter of public record:

Why it happened — three forces, one outcome

The withdrawal isn't fully explained by any single cause. Three forces, each independently sufficient to dampen major-bank SMSF appetite, arrived roughly together:

1. The Productivity Commission flagged prudential risk

The 2018 Productivity Commission inquiry into superannuation ("Superannuation: Assessing Efficiency and Competitiveness") raised explicit concerns about Limited Recourse Borrowing Arrangements (LRBAs) on small SMSF balances. The Commission recommended (gently, with hedging) that prudential treatment of LRBAs on funds under $200,000 deserved scrutiny. The recommendation didn't become law, but it landed in the same news cycle as several broader concerns about SMSF establishment quality — adviser conflicts, suitability gaps for small balances, the rise in low-balance SMSF establishments through aggressive marketing.

For major-bank credit committees, the signal was unambiguous: regulators and policy-makers were paying attention to SMSF lending and were asking whether the prudential framework was appropriate. The cost of a future regulatory tightening — capital adjustments, additional reporting, restrictive APRA standards — sat heavier on a major-bank balance sheet than on a non-bank specialist that was already pricing for narrower regulatory headroom.

2. APRA's targeted reviews tightened residential investor lending broadly

The 2017-2019 period saw APRA progressively tighten residential investor lending across the entire Australian market — interest-only caps, investor loan growth limits, more conservative serviceability assessment. SMSF residential lending sat at the intersection of two segments APRA was constraining (investor + non-standard structure). Even when LRBA-specific rules weren't tightened directly, the broader environment made the segment less attractive to write.

3. The commercial maths got worse

SMSF LRBAs are operationally complex. The custodian trust structure, the s67A documentation, the limited-recourse default workflow, the LMI position (in practice, LMI generally isn't written on LRBA, so lenders carry the entire LVR risk) — none of this is easy. The all-in cost to a major bank of writing a $600k SMSF LRBA is materially higher than writing a $600k owner-occupied refinance. The pricing premium SMSF lending traditionally commanded was being eroded by competition; by 2018, the margin available on SMSF residential wasn't compensating major banks for the operational overhead and policy risk.

Non-bank specialists could carry the operational complexity more efficiently because they were structurally smaller, more concentrated on the file shape, and willing to price the loan at a premium that reflected the cost properly.

The major banks didn't withdraw because they didn't believe in SMSF lending. They withdrew because the segment had become a place where the operational and regulatory cost of writing the loan exceeded the margin available to do it at major-bank pricing.

What it looks like now

Six years on, the residential SMSF panel is structurally different from any other slice of Australian mortgage lending. The active lenders are:

• Prime tier (mutual / retained major): AMP Bank, Bank of Melbourne (Westpac retained), MyState, Heritage Bank, Bank Australia, Credit Union SA, Queensland Country Bank. Rates roughly 6.80-7.40% in mid-2026.
• Specialist tier (non-bank): La Trobe Financial, Liberty, Pepper Money, Resimac. Rates roughly 7.20-8.20%.
• Commercial SMSF (different panel): CBA, Macquarie, Suncorp, Newcastle Permanent still write commercial SMSF (business real property) — they exited residential, not commercial. The two sub-segments behave very differently.

The pricing gap to mainstream residential lending is real — typically 1.00-1.80% above the cheapest owner-occupied rate for an equivalent borrower. On a $700k LRBA, that's roughly $7,000-$13,000 per year in additional interest. That spread is the durable consequence of the 2018-2020 withdrawal: the cost of structurally narrower competition.

Why it still matters today

The retail consequence of the withdrawal is that today's SMSF property strategy involves a meaningfully different lender list than it did pre-2018, with materially different pricing and policy. Three practical implications for trustees and their advisers:

1. Lender choice is narrower and matters more. With 10-12 active residential SMSF lenders instead of 20-25, the right routing matters disproportionately. Getting the wrong lender — one whose minimum balance, liquid-asset buffer or contribution-funded servicing policy doesn't match the fund — can mean the file gets declined outright, not just priced worse.
2. Pricing won't return to pre-2018 levels. The withdrawal isn't structurally reversing. The major-bank operational and regulatory cost calculus that drove the exits hasn't changed materially. Borrowers should price strategy assuming the current panel and its premium, not expecting major-bank competition to compress it.
3. Refinance arithmetic has changed. SMSF LRBA refinances are more constrained than mainstream residential refinances. There are fewer lenders willing to take an existing LRBA off another lender's book; the savings achievable from refinancing are often smaller because most active SMSF lenders price in a similar band. The cost-benefit on refinance is different from the equivalent owner-occupied calculation.

The structural lesson

The SMSF withdrawal is instructive beyond the segment itself. It's an example of how Australian financial market structure shifts: not via a single regulator announcement that gets reported on, but as a cluster of credit-committee decisions made over 18 months in response to overlapping prudential signals, regulatory direction, and commercial cost recalculation. The lenders that exited didn't announce that they were exiting in a single press release. They announced product withdrawals, broker channel changes, criteria tightening — each individually justifiable on its own commercial logic — and the cumulative effect was that the segment changed shape.

This pattern repeats. Industry observers track lender-by-lender announcements; the underlying structural shift only becomes visible in retrospect, when the cluster of decisions is read as a single story rather than separate ones. The panel page on this site is designed to make that pattern visible earlier — every detected policy change is logged, so the cumulative direction shows up before the next "the banks have left" story does.