Brokers Wrote 72% of BOQ’s FY26 Lending and Hold 58% of Its Book. The CFO Says That Reliance Will Moderate — and the Digital Home Loan Already Scores +76 NPS Through Brokers

1. What BOQ actually reported

For the full year ended 31 August 2026, BOQ Group reported statutory net profit after tax of $250 million — up 88 per cent on FY25, which the bank attributes primarily to the non-repeat of a goodwill impairment in the prior year. Cash earnings after tax excluding notable items were $358 million, down 7 per cent. Net interest margin lifted 6 basis points to 1.70 per cent, and the CET1 ratio closed at 11.64 per cent, above the top of the bank’s own 10.25–10.75 per cent management target range.

The mortgage line is the one that will get repeated in aggregator briefings. Home lending contracted $4,633 million, or 8 per cent, which BOQ attributes to lower settlement volumes and higher portfolio run-off “in a period of significant transformation activity”. Commercial lending went the other way, growing $2,062 million or 16 per cent, with the bank naming healthcare, agriculture and diversified businesses as the segments doing the work. Customer deposits contracted $3.8 billion, or 6 per cent, which BOQ ties to a lower funding requirement as housing volumes fell.

There are signs of a floor. BOQ reported housing application volumes of $5.2 billion in the fourth quarter — its strongest quarter in two years — and said that in the second half both application volumes and portfolio run-off improved, “providing momentum for the planned return to home lending growth in FY27”.

2. The channel numbers underneath the headline

BOQ’s housing metrics page carries two figures that frame everything else. Broker share of the existing portfolio: 58 per cent. Broker share of FY26 flow: 72 per cent. In other words, the channel’s share of what BOQ wrote last year was materially higher than its share of what BOQ already held — broker origination is not a legacy position at this bank, it is the current one.

The portfolio-by-channel split in the same presentation shows where that sits by brand: BOQ-branded broker loans at 47 per cent of the housing book, BOQ proprietary at 25 per cent, ME broker at 11 per cent, Virgin Money at 8 per cent, BOQ Specialist at 6 per cent and ME proprietary at 3 per cent. The two broker slices add to the 58 per cent the bank reports.

Then there is the number that complicates the narrative. In a footnote to its customer-experience page, BOQ discloses that the August 2026 spot Net Promoter Score for its Digital Home Loan through the broker channel was +76. The bank’s newest mortgage product, delivered through brokers, is scoring at the top end of anything you will see disclosed in Australian retail banking. The channel is not the problem the bank is solving.

3. Two BOQ brands, two very different propositions

If you have been treating “BOQ Group” as one entry on your panel, the result is a prompt to stop. The presentation footnotes the position plainly: Virgin Money stopped accepting home loan applications from new customers through the broker channel from 1 September 2023, and BOQ-branded home loans stopped accepting new-customer broker applications from 1 September 2024. ME is the group’s retail broker brand.

The application volumes show the consequence. ME took 11,100 housing applications in 2H25, dipped to 9,700 in 1H26 during the platform migration, then recovered to 11,000 in 2H26. BOQ-branded applications sat at 2,700, then 2,800, then 2,600 across the same halves. Group totals moved 14,400, 13,100, 14,300. Put simply: the group’s application recovery in the second half was an ME recovery.

The brand-level lending growth tells the same story from the balance sheet side. Within the retail bank, ME grew $1,279 million in FY26 while the BOQ brand fell $3,572 million and Virgin Money fell $903 million. BOQ’s own commentary says the contraction reflects lower settlements and higher run-off from the closure of legacy channels and the move to the digital platform, “partly offset by ME broker growth”.

The accreditation footprint is the detail that should make principals look at their own panel hygiene. At 31 August 2026, BOQ reported 10,799 brokers accredited with BOQ and 15,640 accredited with ME. The BOQ number is attached to a brand that has not accepted new-to-bank broker applications for two years. If your CRM still lists BOQ as a live new-business option, that is a stale field, not a lender relationship.

4. What “less broker reliance” does and does not mean

The phrase that drove the trade-press coverage came from the results call. According to a transcript of the call published by Investing.com, chief financial officer Racheal Kellaway said the group had “been effectively relying much more on broker than we see going forward”, describing the current mix as “heavily broker, as you would expect for us, given we are just launching the digital prop channel”.

Read the rest of what she said in the same exchange, though, and it is a statement about arithmetic rather than intent. “But look, customers are still tending to like to use brokers for lots of different obvious reasons,” Kellaway said, per the same transcript. “So we expect that to remain a very important channel for us.” Managing director and chief executive Rod Finch was quoted in the same transcript saying “we recognize that brokers are a key factor in this market” and that customers “choose that through the cycle”, with ME as the brand the group takes to market — while “increasingly, we are thinking really about those first-party channels as well”. He described the outcome as “a mix across both first and third parties”.

The mechanism is straightforward. If a bank adds a proprietary digital origination channel from a near-standing start, the broker share of its flow falls even if broker volume holds or grows. A share falling from 72 per cent is not the same as a channel shrinking. Brokers reading this as a retreat from the channel will misread both the numbers and the bank’s stated plan.

What the FY27 outlook does say is specific and worth quoting exactly: the group “has planned for a sustainable return to growth in home lending in FY27”, and “this will be enabled by expanding the Group’s digital mortgage to additional channels and customer segments”. BOQ does not say in the released materials which additional channels, or on what timetable. That is the open question for brokers, and it is a fair one to put to your BDM.

5. The credit settings worth noting

The presentation discloses the group’s serviceability approach in one line: repayment ability is assessed on the higher of the minimum floor rate, “which is currently 5.75%”, or the actual customer interest rate plus a 3 per cent buffer. Confirm the live position with the lender before you rely on it — floor rates move — but it is a useful reference point when you are sizing capacity across a panel.

The flow-versus-portfolio comparison also reveals what BOQ has been writing. FY26 flow ran at 35 per cent investor against 30 per cent in the portfolio; 86.4 per cent PAYG against 81.9 per cent; 17 per cent interest only (excluding construction) against 10 per cent. Only 2.8 per cent of FY26 flow carried a debt-to-income ratio at or above six times, 4.7 per cent was written above 80 per cent LVR and 0.7 per cent above 90 per cent. Fixed-rate flow was 6 per cent.

That is a tightly drawn book: salaried, low-LVR, low-DTI, with an investor and interest-only tilt. If your pipeline is weighted toward self-employed borrowers, higher-LVR first-home buyers or stretched DTI scenarios, the shape of that flow tells you something about where this group is competing and where it is not.

On arrears, the housing portfolio reported 104 basis points of loans 90 days or more past due. The commercial portfolio saw 30-day arrears fall 20 basis points on FY25 with a 5 basis point increase in 90-day arrears, and BOQ says it observed no material deterioration in credit quality while continuing to monitor property and construction closely.

One technical note from the same presentation is worth carrying into any market-share conversation this month: BOQ states that growth rates and system multiples were adjusted to add back approximately $15 billion of residential mortgages that left the APRA ADI population in August 2026 following a major bank portfolio sale to a non-ADI. If you see odd system-growth numbers for August, that is part of the explanation. On BOQ’s own like-for-like basis, ME grew at 0.7 times system while BOQ-branded and group growth were negative.

6. What to review this week

7. What to watch next

Three things will tell you how this actually lands. First, whether BOQ names brokers among the “additional channels” for its digital mortgage, and whether that arrives under the BOQ brand or stays with ME. Second, whether ME’s application run-rate holds above 11,000 a half now that the migration is behind it — that is the cleanest available read on whether the group’s broker volumes are growing while its broker share falls. Third, whether the 1H27 margin uplift the bank has guided to survives the retention discounting pressure the CFO flagged as rising again in recent months.

For now, the practical position is unglamorous and specific: BOQ Group remains a broker-led lender that has split its brands by channel, and the usable entry for new business is ME. The strategic headline is about proprietary distribution. The operational fact on your desk is a panel row that needs updating.