Mortgage broker vs going direct to a lender: the economics, the duty, and when each wins

About 70% of new home loans in Australia are now written through a mortgage broker, up from roughly half a decade ago. The shift is rational on average and irrational on the margin: brokers usually save borrowers money on the headline rate but earn lender-paid commissions that bend their incentives in ways the upfront conversation rarely surfaces. A borrower who understands the economics of the channel can extract most of the upside and price-in the rest.

The choice is not broker-good versus bank-bad. It is a question of which channel suits a particular borrower profile, what the broker is actually being paid to do, and which lender ends up on the application either way. Plenty of broker-written loans land at one of the big four. Plenty of direct-written loans are competitive without a third party in the loop. The right frame is which path is likely to produce the better-priced loan for your file, with conduct rules you can verify.

What a mortgage broker actually does

A mortgage broker is a credit-licensed intermediary who sources loan options across a panel of lenders, packages the application, and manages the file through to settlement. The work is mostly file preparation: payslips, tax returns, bank statements, liability statements, and the lender's servicing calculator filled in correctly. That packaging is non-trivial. A borrower with bonus income, salary sacrifice, or rental income on an existing property can have wildly different serviceability outcomes at two different lenders, and the broker's job is to find the one that says yes at the best rate.

Top brokers carry 25 to 40 lenders on their accredited panel, including the big four, the second-tier banks like Macquarie and Bank of Queensland, the non-banks, and a long tail of customer-owned mutuals. Direct-to-lender locks the borrower into one lender's product set, one credit policy, and one servicing model. If that lender says no, the next step is starting over somewhere else.

How brokers are paid

Brokers are paid by the lender, not the borrower. The commission structure has been steady for a decade or so.

• Upfront commission: roughly 0.65% of the funded loan balance, paid by the lender on settlement. A $620,000 loan pays the broker around $4,030 upfront.
• Trail commission: 0.15% to 0.20% per year on the outstanding loan balance, paid for as long as the loan stays on the lender's book. The same $620,000 loan pays roughly $930 to $1,240 in the first year of trail, declining slightly each year as the principal amortises.
• Clawback: if the loan is refinanced or paid out within the first 18 to 24 months, the lender claws part of the upfront commission back from the broker. Some brokers pass the clawback on to the borrower contractually; many absorb it.

The borrower's direct cost for the broker's service is $0. The lender absorbs the commission as the cost of customer acquisition. The harder question is whether the loan the broker steered you to is better than the loan you could have negotiated directly.

Best Interests Duty and what changed in 2021

Best Interests Duty, or BID, was introduced in January 2021 under amendments to the National Consumer Credit Protection Act 2009. The duty is straightforward in principle: a broker must act in the borrower's best interest, and must be able to evidence that on file. ASIC enforces it. The practical effect is that a broker cannot recommend a loan just because the commission is higher; the file must show how the recommendation aligns with the borrower's stated goals, comparing real alternatives.

BID does not apply to direct lender staff. A bank's own home-loan specialist is bound by the responsible lending obligations of the NCCP and by the bank's internal conduct policy, but not by a personal best-interests obligation to recommend a competitor's product. That asymmetry is the strongest argument for the broker channel for borrowers who do not already have a preferred lender in mind.

Where the broker incentive bends

BID raised the floor on broker conduct. It did not eliminate the incentive structure the commissions create. Two pressures remain.

• Trail commission rewards inertia. A broker who recommends a refinance is voluntarily shrinking their own trail income on the original loan. Most brokers run the refinance anyway when the numbers warrant it, especially because a refinance generates a fresh upfront on the new loan. The borrowers most at risk are those who never call their broker back, on loans where a refinance would have helped but was never raised.
• Upfront scales with loan size. A larger loan, including a higher-LVR product with lender's mortgage insurance, pays a larger upfront. There is a documented incentive to steer marginal borrowers into bigger loans than they would otherwise take. BID is meant to police this; the documentation requirement helps, but the structural pressure is still there.

When a broker wins

The broker channel is materially stronger for some borrower profiles.

• Complex serviceability: self-employed borrowers, casuals, contractors, anyone with multiple income streams, anyone with HECS, child-support, or shared-care complications. Different lenders treat the same income wildly differently. A good broker knows which lender will count 100% of bonus income versus 80%, which lender will accept two years of self-employed returns versus one, and which lender uses actual rent versus a 70% shaded figure.
• Wanting to compare 20+ lenders without 20+ pre-approvals. Every direct pre-approval is a credit enquiry. A broker runs servicing in lender calculators and only submits to the one that wins, protecting the borrower's credit file.
• First-home buyers who would otherwise default to the big four out of inertia. The broker channel routinely surfaces second-tier and non-bank options 25 to 60 basis points cheaper than the big-four sticker rate. See the First Home Guarantee scheme for how scheme eligibility narrows the panel further.
• Investors with cross-collateralisation questions or portfolio re-financing across multiple properties. Direct staff are often not trained for this; specialist brokers structure it weekly.

When going direct wins

Direct-to-lender is not the inferior channel by default. It wins in specific cases.

• Existing customers with a strong relationship at a competitive lender. Big-four retention desks can match or undercut the broker channel for customers they want to keep, especially on owner-occupier loans with strong serviceability.
• Simple PAYG borrowers buying an owner-occupier with 20%+ deposit. The loan is vanilla, the credit decision is fast, and direct customer-retention pricing is often as good as the broker channel. The broker still earns a commission; the direct path saves a few weeks of cycle time.
• Borrowers who want single-point accountability. A direct loan has one party to call when something breaks at settlement. A broker loan has two: the broker and the lender, and they can point at each other.
• Specific products only available direct. Some bank-owned offset structures, professional packages, or private-bank tiers do not flow through broker channels at all.

How to vet a broker

Five questions, asked before the second meeting, separate a competent broker from a mediocre one.

• FBAA or MFAA membership. One of the two industry bodies is the licence-and-conduct baseline. Verify the membership directly on the association's register rather than trusting a logo on a website.
• Lender concentration. Ask which lenders they wrote more than five loans with in the previous twelve months. If 70% of the file load went to one lender, the panel breadth is theoretical rather than real.
• Remuneration disclosure. BID requires the broker to disclose how they are paid. A clear explanation of upfront, trail, and clawback, with the actual numbers for your loan, is the answer you want. A hand-wave is a signal.
• Refinance trail policy. Some brokers keep trail commission on refinances they recommend. Some do not. Ask, and prefer the answer that aligns their incentive with refinancing when the numbers warrant it.
• Recent CDR data. Brokers with serious tech stacks pull Consumer Data Right transaction feeds to model serviceability in detail. Ask whether they do, and how they use it.

A worked example

A PAYG borrower on $80,000 a year, refinancing a $620,000 owner-occupier loan with 22 years remaining, principal and interest. The current lender offers a 0.20% loyalty discount on its 6.30% standard variable, bringing the effective rate to 6.10%. The broker sources a competitor at 5.85%.

The headline saving is 25 basis points on $620,000, or roughly $1,550 in year-one interest. The full saving is slightly larger because lower interest means faster principal amortisation, but the first-year arithmetic is a reasonable approximation. Over five years, the interest saving is in the range of $7,000 to $7,500, depending on the exact balance trajectory.

The broker earns roughly $4,030 upfront on the new loan and $930 to $1,240 in first-year trail. The borrower pays none of that directly; the new lender absorbs it as an acquisition cost. The borrower's gross saving is the rate differential. The broker's gross earnings sit on top, paid by the lender, drawn from the same rate margin the lender is willing to give up to win the loan. Model your own version on the loan comparison calculator with both rate scenarios side by side, and double-check the all-in cost using the comparison rate tool so the fees do not eat the rate differential.

What this means for the borrower

Three useful conclusions sit at the end of all this.

• For most borrowers without an existing strong banking relationship, the broker channel is the higher-expected value path. Panel breadth, BID, and the lender-paid commission structure mean the borrower captures most of the price competition with no out-of-pocket fee.
• For borrowers with a strong relationship and a competitive direct offer in hand, the direct channel can match or beat the broker channel. Ask the broker to compete with the direct offer; the price discovery itself is informative.
• The conduct rules are not optional. BID is enforceable. A broker who cannot articulate, with numbers, how their recommendation is in your best interest is failing the basic test of the licence. Walk.

The companion piece on switching home loans covers the mechanics once the lender choice is made. The APRA serviceability buffer explainer is the regulatory layer that sits between the rate on the application and the borrowing capacity any lender will approve. Reading both alongside this article gives a borrower a complete view of the channel decision, the rate decision, and the capacity decision in one sitting.