High-DTI Lending Is 5.6% Against APRA’s 20% Cap. The Decline on Your Six-Times File Is a Lender Call, Not a Regulatory One

What APRA actually published this week

APRA’s System Risk Outlook is a twice-yearly information paper on risks facing the financial system. The November 2026 edition landed on 12 November. Its headline findings are not, on their face, broker news: household debt sits at around 180 per cent of disposable income, which APRA describes as high by both historical and international standards, and the paper warns that “high household indebtedness could amplify a more severe downturn.”

APRA Chair John Lonsdale framed the overall position this way in the accompanying media release: “Australia’s financial system remains resilient, but the risk environment is becoming less forgiving.” He also said the system “is well placed to withstand this turbulent period.”

Buried in the lending section are the two sentences that matter for the broker channel. The first: “Maintaining sound lending standards is important as lenders compete strongly for borrowers in a slowing market.” The second is APRA’s confirmation that the debt-to-income limits introduced in February 2026 remain a safeguard against a renewed build-up of riskier lending, and that APRA regularly reviews its macroprudential settings.

Read those together and a signal emerges. Volumes are falling. Competition for the borrowers who remain is intensifying. APRA has noticed, and it is explicitly keeping the brake on while that plays out. APRA also notes that it and the other Council of Financial Regulators agencies are monitoring whether competition — from banks and from non-bank lenders — is weakening standards, while assessing that riskier lending currently remains contained.

The cap, precisely

It is worth restating what the limit actually is, because it is routinely described loosely in client conversations.

Under APRA’s macroprudential framework, a residential mortgage with a debt-to-income ratio of six or more may make up no more than 20 per cent of an authorised deposit-taking institution’s new mortgage lending. The detail that matters in practice:

• The 20 per cent limit applies to owner-occupier and investor portfolios separately, and is measured quarterly.
• It applies to new lending only. Existing borrowers are not captured.
• It applies to ADIs. Smaller institutions get proportionate treatment, including a four-quarter rolling measurement. Non-ADI lenders are not subject to the limit at all.
• Two categories are exempt: bridging loans for owner-occupiers, and loans for the purchase or construction of new dwellings.
• It commenced on 1 February 2026.

Note what DTI captures. As Westpac Group Chief Economist Luci Ellis observed when the limits were announced, debt-to-income counts total debt — including HECS-HELP balances and buy-now-pay-later facilities — which is why the measure bites hardest on investors, younger borrowers and clients carrying several facilities.

The gap between the ceiling and the floor

Here is where the published data gets interesting. APRA’s quarterly ADI property exposures statistics for the June 2026 quarter — the most recent available — show:

• DTI of six or more: 5.6 per cent of all new loans, up from 5.5 per cent a year earlier.
• New owner-occupied lending at DTI six-plus: 3.7 per cent, essentially flat year on year.
• New investment lending at DTI six-plus: 8.9 per cent, up from 8.7 per cent.

Against a 20 per cent ceiling, the owner-occupier bucket is under a fifth full. The investor bucket — the one that was always going to bind first, and where investors now make up 35.6 per cent of new loans funded — is under half full.

This was the picture Canstar Data Insights Director Sally Tindall described when the limits were first announced, when DTI six-plus lending was “just 5.5% of all new lending.” Her caveat then is the one to hold onto now: this type of lending “can quickly climb over 20% in a low-rate environment.” It has not climbed. On the latest published data, it has barely moved.

One honest qualification belongs here, and it is the hinge of everything that follows. These are industry-wide figures. APRA does not publish each ADI’s individual DTI share. It is entirely possible for a particular lender to be sitting much closer to its own limit than the aggregate suggests, particularly on the investor side. What the aggregate tells you is that a system-wide shortage of high-DTI capacity is not what is happening.

So why was your file declined?

Because, on the available evidence, the regulatory cap is unlikely to be what is binding. Lenders are choosing to sit well below it.

Brokers began reporting this within weeks of commencement. Speaking to Australian Broker in March 2026, Eventus Financial founder and mortgage broker Alex Veljancevski said that “some banks are now limiting their exposure to higher debt-to-income loans more actively,” and, more pointedly, that “some have already quietly lowered their own DTI limits internally.”

That is a credit appetite decision, not a compliance obligation. And the distinction is not academic for three reasons.

It changes the client conversation. “The regulator won’t allow it” is both inaccurate and strategically unhelpful, because it implies nothing can be done and no other lender will differ. “This lender’s internal policy sits below the regulatory ceiling, and lenders differ on where they set it” is accurate, and it keeps the file alive.

It changes lender selection. If internal DTI tolerance varies lender by lender and is not published, then it becomes a material input into placement that you can only get by asking. That is a BDM conversation, and the question is not “are you near the APRA cap?” — almost no one is. The question is “what is your internal DTI ceiling, is it different for owner-occupier and investor, and do you write exceptions above it?”

It changes what you record. Where a client’s circumstances mean a higher-DTI structure is genuinely appropriate, the reasoning that supports your recommendation sits in your file, not in the lender’s policy. Under the Best Interests Duty, the recommendation has to stand on the client’s circumstances and objectives — the fact that a particular lender happened to have appetite is not, by itself, the reason a loan is in a client’s best interests.

The two exemptions that are being left on the table

The most under-used fact in this entire framework is that the cap does not apply to every high-DTI loan.

Bridging loans for owner-occupiers are exempt. Loans for the purchase or construction of new dwellings are exempt. A DTI six-plus file in either category does not consume a lender’s limited bucket.

For brokers, that has real placement consequences. The upsizing owner-occupier who needs a bridge, and the client buying or building new, are structurally better positioned on high-DTI files than the equivalent client buying established stock — not because their risk profile is better, but because of how the instrument is drafted. If you are working a client whose DTI is going to land at or above six, the question of whether the transaction is a new build or an owner-occupier bridge is no longer just a product question. It is a capacity question.

One caution: an exemption from APRA’s portfolio limit is not an exemption from the lender’s own credit policy or serviceability assessment, and it is certainly not an exemption from responsible lending obligations under the National Credit Act. It removes one constraint, not all of them.

The non-bank question, handled properly

Non-ADI lenders are outside the limit entirely. That makes them the structural home for a genuinely high-DTI file that no ADI will write.

It does not make them a default. Veljancevski’s warning when the cap commenced was blunt: non-banks are not covered by the cap, but may charge higher rates or impose different conditions, so “it’s not a straight swap.”

Two things follow. First, a move to a non-bank on a high-DTI file is a recommendation that has to be justified on its merits for that client — pricing, conditions, exit, and the client’s own objectives — and documented as such. Availability is not a reason. Second, brokers should note that APRA used this week’s paper to flag that it and the Council of Financial Regulators agencies are watching whether competition from non-bank lenders is eroding lending standards. That is not an allegation of anything against any lender, and APRA’s assessment is that riskier lending remains contained. But it does mean the segment carrying the uncapped high-DTI business is a segment the regulators have said they are watching, and appetite there is worth treating as a live variable rather than a fixed feature of your panel.

What this week removed: the hope of loosening

There is a reasonable instinct in a downturn that macroprudential restraint gets eased. Prices are falling. Lending volumes have slumped. Arrears, while rising, remain low — APRA cites RBA estimates that fewer than 2 per cent of variable-rate owner-occupier borrowers face a cash flow shortfall, and fewer than 1 per cent of all borrowers are in negative equity. On that picture, a broker could be forgiven for assuming the DTI brake comes off and borrowing capacity comes back.

APRA has just said otherwise, in the same document that confirmed the price falls. The limits remain a safeguard. The reasoning is visible in the paper: the risk APRA is managing is not today’s arrears, it is lenders competing harder for a shrinking pool of borrowers. A falling market is the argument for keeping the brake on, not against it.

MFAA CEO Anja Pannek read the original announcement as “deliberately pre-emptive rather than reactive.” The November paper is consistent with that reading. Brokers building a 2027 pipeline on the assumption that capacity constraints ease should plan on the opposite.

What to review this week

1. Pull your declines. Identify every application knocked back in the last six months where DTI was at or above six. Tag which were declined on lender DTI policy versus serviceability, credit history or security. Most brokers assume the first category is larger than it is.
2. Ask four lenders the same question. What is your internal DTI ceiling, does it differ between owner-occupier and investor, do you write exceptions above it, and who approves them? You will not find this in a product guide.
3. Flag the exempt transactions. In your current pipeline, mark every owner-occupier bridging file and every new-build or construction purchase. These do not consume a lender’s high-DTI capacity.
4. Audit debt hygiene before lodgement. Because DTI counts total debt, a closed BNPL facility or a cleared small personal loan can move a client across the six-times line. That is a pre-lodgement conversation, not a post-decline one.
5. Fix the script. Replace “the regulator caps this” with an accurate explanation of lender-level policy variation. The first version loses the client. The second keeps the file in play.
6. Check your file notes on high-DTI recommendations. Where you have recommended a lender because they had appetite, make sure the file also records why that loan suited the client’s circumstances and objectives.

What to watch next

Three things. APRA’s next quarterly property exposures release, which will show whether the investor DTI six-plus share is climbing from 8.9 per cent or holding. Any movement in the cash rate, given Tindall’s warning that high-DTI lending climbs quickly in a low-rate environment — that is the condition under which the cap stops being theoretical. And APRA’s own statement that it regularly reviews its macroprudential settings, which cuts both ways: the 20 per cent figure is not permanent in either direction.

The practical takeaway

A regulatory cap with three-quarters of its headroom unused is not what is costing you high-DTI deals. Individual lender appetite is — appetite that varies between lenders, is not published anywhere, and that APRA has just given the industry fresh reason to keep conservative.

That is a worse problem than a hard cap in one respect: you cannot look it up. But it is a better problem in the respect that matters commercially, because a constraint that varies between lenders is a constraint that rewards the broker who knows where it sits. Right now, most do not know, because most have been told the ceiling and assumed it was the floor.