Key implication: 2026 will be a market‑share redistribution year. Margins stay tight, distribution power consolidates around brokers and digital channels, and the winners will fuse open‑data plumbing with disciplined AI to reduce days‑to‑decision, lift pull‑through, and keep customers from refinancing away.

Macro reality: Lower rates may not equal easy growth

The Reserve Bank of Australia’s February 2025 Statement on Monetary Policy noted forecasts conditioned on market expectations for a gradual easing path, but the growth profile remained subdued relative to pre‑pandemic norms. Deloitte’s 2026 banking outlook highlights a plausible downside in which trade frictions and inflation stickiness slow growth further, keeping funding costs and credit risk prudence elevated. Translation: demand returns unevenly; credit appetite improves, but lenders can’t bank on a broad‑based volume surge.

The upshot for mortgage leaders is a doubles game—volume growth via switching and share capture, margin defended via tighter risk‑based pricing and cost‑to‑serve reductions. As NAB’s Adam Brown observed of 2025, market activity was steady rather than exuberant, a tempo likely to continue into early 2026. In that environment, operational speed and customer effort, not rate alone, become decisive.

Competition and regulation: CDR moves from promise to plumbing

The ACCC’s 2024 paper on strengthening competition in financial services underscored a simple theme: timely reform should reduce switching frictions so consumers and businesses realise the benefits of choice. Australia’s Consumer Data Right (open banking) remains the regulatory backbone. In 2026, practical use beats policy: lenders that embed CDR‑enabled income and expense verification, instant liability checks, and pre‑populated applications will shorten time to unconditional approval and compress abandonment.

Expect intensified watchdog attention on fairness, fees, cross‑selling, and switching impediments. For banks and non‑banks, this shifts the game from marketing horsepower to process engineering: conversion rates and time‑to‑yes become board‑level KPIs, audited as closely as net interest margin. A technical note: consent orchestration and token management for CDR need to be industrial‑grade; poor consent flows will quietly kill adoption.

Technology and AI: From pilots to production—with guardrails

Australia’s AI ecosystem, per a 2025 sector review, shows strong adoption but a gap in commercialisation readiness. For mortgage lenders, that gap manifests as proof‑of‑concept purgatory: clever models that never hit the credit factory floor. The Australian Taxation Office’s AI governance work and the government’s AI Ethics Principles signal the bar: explainability, accountability, and security aren’t optional.

Technical playbook for 2026 production AI in mortgages:

• Credit decisioning: Gradient‑boosted and neural models can enhance behavioural scoring and debt‑serviceability stress testing; keep them within challenger frameworks alongside traditional scorecards, with champion‑challenger rotation and stability monitoring.
• Income and expense analytics: CDR and bank‑feed categorisation models reduce manual assessment; pair with rules‑based overrides for edge cases to satisfy responsible lending and audit queries.
• Servicing intelligence: Large language models, ring‑fenced behind retrieval layers, can summarise case notes and triage broker queries; rate‑limit outputs and implement human‑in‑the‑loop for adverse decisions.
• Model risk governance: Maintain model inventories, bias testing by cohort (age, postcode, employment type), and post‑implementation monitoring windows; document decision rationale retrievably.

The commercial lens: treat AI as an operating system to reduce cycle time, unlock straight‑through processing for simple cases, and free assessors for complex files. Organisations that shift from sporadic pilots to reusable components (data pipelines, consent services, feature stores) will see compounding returns.

Distribution dynamics: Brokers still rule—so make them faster

Australia’s broker channel remains the dominant origination gateway, with steady activity through 2025. In 2026, broker loyalty will tilt to lenders that remove friction: single‑sign‑on portals, pre‑validation of documents via CDR, and live SLA transparency. Lenders should measure and publish real SLA performance, not averages. The operational objective is ruthlessly practical—cut rework, eliminate back‑and‑forth, and surface pricing tiers upfront to avoid late‑stage declines.

For aggregators, the differentiator shifts from panel breadth to tech enablement: integrations that pass structured data (not PDFs), automated compliance pack assembly, and real‑time status APIs. Expect competition to intensify around refinance packages as ABS lending indicators continue to foreground the number of refinanced commitments in seasonally adjusted terms—a reminder that churn, not just first‑home demand, is the growth engine.

Economics of retention: The cheapest deal is the one you don’t lose

With refinancing velocity elevated, the ROI calculus is blunt: it is cheaper to identify and proactively reprice an at‑risk borrower than to win an equivalent new customer at full acquisition cost. A practical retention stack for 2026 looks like this:

• Early‑warning models scanning rate‑expiry cohorts, payment behaviours, and external rate competitiveness.
• Trigger‑based offers delivered via in‑app, email, and broker channel, with consented use of CDR data to tailor offers.
• One‑click execution for internal refinances or product switches, with digital verification and e‑signing.
• Broker‑aligned incentives that reward save‑actions, not just new lodgements, minimising channel conflict.

Boards should ask for a single retention P&L that makes the trade‑offs transparent: foregone margin vs expected lifetime value, churn risk by segment, and operational load on credit and settlements.

Scenario planning for 2026: Three paths and their plays

Base case: Gradual easing, steady housing turnover, continued refi churn. Plays: double down on CDR onboarding, expand risk‑based pricing bands, and automate verification to lift pull‑through.

Upside: Faster‑than‑expected rate cuts spur investor and first‑home activity. Plays: pre‑approved capacity in operations (flex staffing, decisioning automation), dynamic pricing guardrails to protect margin during spikes.

Downside (Deloitte‑flagged risk): Global inflation or tariff shocks keep rates higher for longer; arrears tick up. Plays: tighten serviceability for exposed cohorts, deploy hardship analytics early, and prioritise capital‑light retention over aggressive acquisition.

Execution reality: What it takes to win

Strategy dies in integration. Leaders should mandate:

• Data readiness: Clean income/expense taxonomies, mapped to CDR schemas; consent and identity services centralised.
• Process control: Straight‑through processing for simple files; expert queues for complex; SLA telemetry visible at executive level.
• Risk and ethics: Model governance aligned with AI Ethics Principles; robust documentation for regulatory scrutiny.
• Partner model: Vendor SLAs tied to measurable outcomes (cycle time, abandonment), not just deliverables; broker‑tech co‑design sessions quarterly.

The signal from 2025 is clear: stable activity, fierce competition. In 2026, the market will reward institutions that turn regulation into rails, AI into throughput, and broker experience into a strategic moat.

Sources: The Adviser lender outlook (2026), ABS Lending Indicators (September Quarter 2025), ACCC—Framing the future of financial services (31 Oct 2024), RBA Statement on Monetary Policy (Feb 2025), Deloitte 2026 Banking & Capital Markets Outlook (Oct 2025), Australia’s AI ecosystem report (June 2025), Australian Government AI Ethics Principles.