95% of Macquarie’s New Loans Come Through Brokers. If They’re 40% of Yours, That’s a Single Point of Failure

1. The number the headlines left out

The FY26 result is clean. Home loans $181.3 billion at 31 March 2026, up 28 per cent. Market share approximately 7.1 per cent. Growth driven primarily by lower-LVR and owner-occupier lending. Average LVR at origination 65 per cent; average dynamic LVR 51 per cent. Net interest income rose 7 per cent, and the bank warned that “market dynamics and portfolio mix… continue to drive lower margins”. Group NPAT was $4.8 billion, up 30 per cent. Second-half profit was $3.2 billion, up 93 per cent on the first half. Full-year dividend $7 a share, up 50 cents.

That is a bank that is not under earnings pressure. It is also a bank whose mortgage machine does not run without you. More than 95 per cent of all new Macquarie home loans are originated through brokers. Macquarie’s own 2025 campaign put the industry figure at about 75 per cent of new Australian home loans. The gap is the story. This is not a major that happens to like brokers. It is a broker-originated book with a banking licence.

Brown called 7.1 per cent “a big milestone if you consider the highly competitive nature of the market.” She also said: “What our results show is that it is possible to be profitable lending via the broker channel.” Both are fair. A third reading belongs in the same briefing: a lender that cannot originate without the channel is a lender whose operational decisions — turnaround, policy, BDM coverage, accreditation — land on the channel first.

2. From 0.2 per cent to 7.1 per cent

In 2010, Brown told Australian Broker, Macquarie was 0.2 per cent of the market. By March 2019 the APRA monthly series had the book at $36.1 billion and 2.1 per cent share. Seven years later the same series has it at $173.7 billion — up 380 per cent — and 7.1 per cent of a national market worth $2.46 trillion in March 2026.

Two book figures will appear in the same week of coverage. They are not a contradiction. APRA’s March 2026 ADI statistics put the book at $173.7 billion, up 27.1 per cent year on year. The FY26 result, struck at 31 March 2026, is $181.3 billion, up 28 per cent. Different series, different cut. Use FY for the year just printed. Use APRA to compare with the majors.

On that APRA series: CBA grew 7.1 per cent to $624.5 billion; NAB 5.5 per cent to $347.0 billion; Westpac 5.3 per cent to $509.0 billion; ANZ 3.7 per cent to $324.2 billion. The market was up 6.9 per cent. Macquarie added $3.6 billion in March alone — about 2.1 per cent in a month, against 0.4 to 0.5 per cent for the majors. The big four’s combined share is 73.3 per cent, down from 78.9 per cent in March 2019. That shift did not sprinkle evenly across a dozen regionals. A large part of it accumulated in one broker-originated book.

3. A broker bank, almost entirely

Broker Pulse’s 2026 Third-Party Lending Report surveyed more than 1,200 residential brokers between 17 February and 30 April 2026. Macquarie had the highest broker usage in the past 12 months, at 70 per cent, for the second year running. It topped all five categories and won Non-major bank (Large).

Product policy is the top factor in choosing a lender, named by 93 per cent of brokers. Brokers put files where the policy fitted, the credit decisioning was clear and the turnaround held. Macquarie has spent years making that the rational choice: Bank of the Year in the 2025 MPA Brokers on Banks survey for a fourth consecutive year; MFAA Major Lender of the Year for a sixth. Brown said the bank had doubled down on BDM and credit assessment teams and digital experience, and remained committed to turnaround times, clear credit decisioning and responsive support. That is why the book looks like this. It is also why the concentration risk now sits on your side of the desk. If 70 per cent of the channel already uses the same lender, a policy rewrite or a TAT blowout does not hit a niche. It hits the default path.

4. The book is well-secured. The panel is not

Macquarie’s credit settings are not the risk in this piece. Average LVR at origination 65 per cent; average dynamic LVR 51 per cent; growth driven primarily by lower-LVR and owner-occupier lending. This is a well-secured book talking.

The compositional wrinkle is investor share. On the March APRA cut, Macquarie’s investor book was $67.3 billion against $106.4 billion owner-occupier — roughly 38.7 per cent investor. Nationally, investor balances were $800.5 billion, or 32.5 per cent of the $2.46 trillion market, with owner-occupier at $1.66 trillion. Heavier than the majors. Heavier is not reckless. It is a mix worth knowing when you lodge the next file.

The backdrop is worth dating, not extrapolating. APRA’s DTI limits on new lending at DTI of 6 times and above activated on 1 February 2026. When MPA published the March cut on 1 May, the RBA had hiked twice in 2026 and inflation sat at 4.6 per cent. Moody’s had said a further rise in the cash rate to 4.6 per cent would push national housing affordability above 31 per cent of average disposable income. That is May reporting, not a forecast, and not a claim Macquarie is about to tighten. A conservative book can still change policy, stretch turnaround, thin BDM coverage or pause accreditations. Those are service events. They do not appear in an LVR.

5. What concentration actually costs

Four failure modes. None require a prediction that Macquarie will turn.

Turnaround. If 40 per cent of last quarter sat with one lender, a TAT blowout is 40 per cent of your clients waiting. The rest of the panel does not absorb that automatically. You absorb it — in phone time, in referring-partner explanations, and in files that age past the unconditional date.

Policy change. Product policy is why 93 per cent of brokers choose a lender. If Macquarie is your default for lower-LVR owner-occupier, a rewrite is a gap you discover on a live file, not a niche product going away.

BDM coverage. Brown said the bank doubled down on BDM and credit assessment teams. That is the current investment, not a covenant. A book that grew 380 per cent in seven years will keep testing the people who pick up the phone.

Accreditation. A pause or a product stop is admin at 12 per cent of settlements. At 40 per cent it is a revenue event, and referring partners notice first.

The bank flagged the other pressure: net interest income up 7 per cent, with a warning that market dynamics and portfolio mix continue to drive lower margins. Margin rebuild is a management decision. It can arrive as pricing, as policy, or as quieter service rationing. You will not get a memo titled “we are rationing service”. You will get slower decisions and a fuller BDM.

6. Four bands for last quarter

If you cannot name the share, that is the finding. Pull last quarter’s settlements. Split them by lender. Dollar share pays the bills; file-count share is how many client relationships sit on the same path.

Under 15 per cent. You have a panel. Keep a second lender warm on the same borrower types so the muscle memory is there.

Fifteen to 30 per cent. A common, rational weighting. The risk is substitution speed. Lodge one comparable file this week with your next-best lender and time the decisioning.

Thirty to 50 per cent. The band the headline is written for. Map every live Macquarie file to a named fallback — lender, BDM, accreditation.

Over 50 per cent. That is not a panel. Cap new Macquarie lodgements this month at a share you can defend, and rebuild two alternative pathways.

The tool below walks the same four bands. The arithmetic takes twenty minutes. The conversation with yourself takes longer.

7. Three actions this week

1. Count last quarter. Settlements by lender, in dollars and in file count. Manage to the dollar share. Do not ignore the file count — that is how many conversations you will have if turnaround slips.
2. Name a fallback for every live Macquarie scenario. Not “another bank”. A named lender, a named BDM, a confirmed accreditation, and one recent file that proves the path still works.
3. Lodge one non-Macquarie file this week on a scenario you would normally send to Macquarie. You are buying current information about turnaround and credit decisioning. That is research, not a political gesture.

Brown is right that it is possible to be profitable lending via the broker channel. She is right that brokers are driving the success. The unfinished sentence is that a channel which created a 7.1 per cent bank can also be the channel that discovers it was too dependent on one.