The headline is simple: Australia’s new mortgage flows are back, rising 18.9% in the September 2025 quarter versus a year earlier, according to APRA. The implication is not. A market that spent two years rationing risk and fighting for deposits must now scale origination without sacrificing margin or credit quality—just as rate-cut expectations drift into 2026 and household stress indicators remain stubborn.

Demand is back, but the cost of money isn’t

The surge arrives even as major banks push their cash rate-cut forecasts into 2026, with some signalling February as the earliest plausible window. That mix—rising demand with still-high funding costs—typically compresses net interest margins (NIMs) as lenders sharpen front-book pricing to chase share. The August 2025 move by BOQ to drop rates to the market’s floor underscores how hungry lenders have become to seed growth and defend volumes.

Earlier in 2025, industry commentary suggested business lending was outpacing mortgages. APRA’s latest read indicates the pendulum has swung back to housing. The strategic pivot now is converting this cyclical tailwind into economic profit, not just headline growth.

The economics: growth that pays its way

Two variables will decide whether this expansion creates value: acquisition cost and risk-adjusted yield.

• Acquisition cost: With customers shopping aggressively and refinance churn elevated, lenders will shoulder higher front-book discounting and broker commission outlays. Australia’s customer acquisition is heavily digital; the ACCC notes Google maintains “a market share of nearly 94 per cent” of search, concentrating performance-marketing spend and keeping cost-per-acquisition high.
• Risk-adjusted yield: With rate cuts not imminent, repricing lags and deposit competition can cap NIM uplift. The credit side matters: the Australian Financial Security Authority reports excessive borrowing as the leading self-reported cause of personal insolvency (37.3% of debtors), a reminder that volume-led strategies must ride alongside sharper serviceability and early-warning controls.

In short, volume is back, but the value of that volume depends on smarter origination funnels, precise pricing, and early risk detection. A simple “cut rate, win share” play will hurt later in the cycle.

Distribution advantage: brokers, digital funnels and the Google tax

The battleground is distribution. Broker channels remain pivotal in steering flow, and the marketing stack around them is consolidating. The ACCC’s confirmation of Google’s enduring search dominance keeps paid search costs elevated, punishing undifferentiated acquisition strategies. Lenders and aggregators need to reduce dependence on auctioned clicks by building owned channels—customer referral flywheels, first-party data activation, and pre-approval programmes that lock in intent before rate shopping intensifies.

For challengers, BOQ’s pricing gambit illustrates a broader opportunity: use tactical front-book pricing to wedge into growth segments, while ringfencing margin with surgical retention and granular risk-based pricing at renewal. For majors, the opportunity is to fuse broker excellence with data-driven retention—improving save rates at expiry via proactive, segment-specific offers rather than blanket cuts.

Risk governance meets AI: scale without slippage

The operational challenge is scaling underwriting and fraud controls as applications spike. General-purpose AI now powers faster document processing, anomaly detection and decision support. The Australian Taxation Office’s guidance describes “general purpose or strong AI — an AI system that can be used for a [range of tasks]”, a reminder that off-the-shelf models can supercharge productivity but also broaden risk if not tightly governed.

Australia’s AI Ethics Principles set the baseline for responsible deployment in financial services. Practically, that means:

• Model governance: clear lineage, versioning, and outcome monitoring to detect drift, bias and false acceptances.
• Human-in-the-loop: binding human oversight for exceptions and high-risk cases, with auditable decision trails.
• Data minimisation and quality: strict controls on data usage, and validation routines for income, expense and asset verification.

Early movers can reduce turnaround times materially and lower unit costs per settled loan. But boards should expect APRA and auditors to probe AI explainability, adverse action notices, and fairness testing—especially in low-doc or near-prime segments.

Macro and household resilience: watch the weak joints

The lending upswing is occurring alongside mixed household resilience signals. Insolvency data shows borrowing-related distress remains a leading factor. If cash rate relief is delayed, a portion of recent originations—particularly high DTI and high LVR pockets—will face serviceability pressure. Portfolio managers should intensify early-warning systems (missed payment heuristics, rising expenses, BNPL signals where visible), while hardship operations prepare for elevated case volumes into mid-2026.

From a PESTLE lens: policy risk (serviceability buffers and responsible lending scrutiny), economic uncertainty (rate path slippage), social dynamics (renter-to-buyer shifts as vacancy remains tight), technology (AI and automation), legal obligations (privacy, explainability), and environmental exposures (climate risk and insurability) all sit upstream of credit outcomes. The current growth spurt is, in effect, a live-fire test of post-pandemic risk frameworks.

Competitive playbook: where profit pools move next

Expect product simplification and sharper segmentation. Three moves stand out:

• Segment-led pricing: deploy micro-segment pricing based on channel, property type and risk cohort rather than headline rate wars.
• Flow-to-fund conversion: compress cycle times with pre-populated applications, straight-through verification and conditional approvals that hold for defined windows, reducing fallout from competitor poaching.
• Retention science: elevate save offers from reactive to predictive—identify at-risk cohorts 90 days pre-expiry and present personalised, economically rational offers that preserve NIM.

On the broker side, the edge lies in advisory depth—budget stress testing at higher-for-longer rates, loan structure optimisation, and lender selection tuned to credit policy variation. As volumes swell, time-to-yes and policy clarity will beat a marginally cheaper sticker rate.

Implementation reality: what to change on Monday

For CEOs and CROs, the immediate checklist is operational and measurable:

• Capacity-to-demand match: flex underwriting and settlement capacity to keep SLAs under five business days during peaks.
• AI with guardrails: deploy document classification and fraud models with continuous monitoring; record rationale for adverse decisions.
• First-party data activation: build owned pre-approval journeys; reduce paid-search dependence by 10–20% through referrals and content.
• Risk early-warning: tighten triggers on expense inflation and revolving credit usage; warm-transfer customers into hardship before arrears age.
• Board reporting: split front-book vs back-book NIM, broker vs direct channel economics, and cohort-level delinquency to spotlight profitable growth.

Outlook: growth, then gravity

The likely path through 2026: steady origination momentum into H1, a tentative rate inflection later in the year, and a lagging uptick in hardship claims from stretched cohorts. The strategic prize will accrue to lenders that treat this rebound as a systems upgrade—modernising distribution economics, hardening AI governance, and aligning credit appetite to a slower, stickier rate cycle. Volume is the headline. Quality growth is the story.