Upgraders Hit a Record 46% of AFG’s Lodgements While First Home Buyers Held at 12%. RG 273.91 Puts the Strategy Inside the Recommendation, Not Just the Product

The number that should change how you read your own pipeline

Australian Finance Group released its Mortgage Index for the first quarter of FY2027 on 13 October 2026. The headline figures were soft but unremarkable: $26.8 billion in residential lodgements, down 4.5 per cent on the prior quarter and 12.4 per cent on the same quarter a year earlier, across 36,720 mortgages. The average loan size set a record at $730,145, up 4.5 per cent year on year. The national average loan-to-value ratio sat at 63.1 per cent, down from 63.4 per cent and among the lowest readings in recent years.

The figure worth stopping on is in the buyer-type split. Upgraders accounted for 46 per cent of lodgements — a record in the Index — and AFG reported that monthly activity in September approached 49 per cent. Investors fell to 32 per cent from 34 per cent. First home buyers held at 12 per cent. Refinancers slipped to 15 per cent from 16 per cent.

Put those four numbers next to the same quarter a year earlier and the shape of the change is clearer. In Q1 FY2026, AFG reported investors at 36 per cent of new loans and refinancing at 17 per cent of lodgements, on $30.6 billion across 43,799 mortgages. Total volume has come down roughly 12 per cent year on year, and within a smaller pie, one segment has taken a materially larger slice.

What AFG’s chief executive said in November

Speaking at AFG’s national conference in Melbourne on 10 November, chief executive David Bailey updated those figures, according to a report by The Adviser. He put the upgrader share of AFG’s flow at “closer to 50 per cent these days”, up from around 42 per cent. On first home buyers, he said the share sat at about 12 per cent and that “our data says that hasn’t moved”, adding: “Customers aren’t going to catch a falling knife, especially new home owners.”

Bailey put investor lending at about 30 per cent of flow against what he described as a long-term average of 37 per cent, and refinances at 15 to 16 per cent of volume, which he suggested could lift to 20 to 25 per cent over the next two to three months once the summer lull passes. Those are his characterisations of AFG’s data in November rather than published Index figures, and they sit slightly below the Index on investors and slightly above it on upgraders. Both point the same direction.

The published Index also noted that refinancing is unusually low for this point in the rate cycle, which Bailey attributed to serviceability constraints, rate uncertainty and lenders working harder to retain existing customers. And in the Index commentary he observed that upgraders “are the first to react” when house price conditions turn, while maintaining that “the role of the broker remains critical” in an uncertain market.

The data predates the November rate rise

This is the part that is easy to miss. The quarter the Index measures closed on 30 September. The Reserve Bank’s Monetary Policy Board raised the cash rate target by 25 basis points to 4.85 per cent on 3 November. The November Statement on Monetary Policy reports headline CPI at 3.9 per cent over the year to the September quarter and the trimmed mean at 3.7 per cent, forecasts headline inflation to peak at 4.5 per cent in December 2026, and says headline and underlying inflation are expected to “remain above 3 per cent until end-2027”. Unemployment averaged 4.6 per cent in the September quarter and is forecast at 4.7 per cent by December 2026.

So the record upgrader share was set in a quarter that ended before the most recent increase landed. Whatever the hike does to the segment, the Index has not measured it yet.

There is an early read on sentiment. The Westpac–Melbourne Institute mortgage-holder sentiment measure fell 11.6 per cent in November to 68.1, its lowest level since the series began in 1994, as reported by The Adviser on 12 November. The broader Westpac Consumer Sentiment Index fell 2.7 per cent to 78.2. The survey’s rate expectations index eased slightly to 177.8, with 72 per cent of respondents still expecting mortgage rates to rise over the next 12 months, down from 80 per cent in October. These are survey measures of intent, not settlement data — but upgrading is a discretionary, confidence-led transaction, and upgraders are the segment AFG itself describes as the first to react.

Why concentration is a different problem from a downturn

A falling market is a volume problem. Concentration is a structural one, and brokers tend to underrate it because the monthly numbers look fine right up until they don’t.

Three things follow from half your flow sitting in one segment. The first is correlation: the trigger that stops one upgrader — a soft valuation, a slow sale, a servicing shortfall after a rate move — tends to stop all of them at roughly the same time. Second, there is no obvious replacement engine. First home buyer share has been flat for a long period on AFG’s own data, investors are running below their long-run average, and refinancing is subdued for reasons that are structural rather than seasonal. Third, the upgrader file is the most operationally complex transaction a residential broker writes. It is two property events, usually with two sets of timing, two valuations and a financing strategy that has to bridge the gap between them.

That third point is where the compliance exposure concentrates too.

The upgrader file is a strategy decision before it is a product decision

Brokers tend to think of the best interests duty as a product-selection obligation: did you consider enough lenders, did you pick a sensible one, did you record why. For an upgrader, the product question is downstream of something larger.

ASIC’s Regulatory Guide 273 is explicit that the duty reaches further than the product. RG 273.91 states that you must base all recommendations on the consumer’s individual circumstances, and that this includes the judgements you make about “the strategies, types of credit product, specific credit products and credit providers you investigate” and those you “make recommendations about”. Strategies appear first in that list, and separately from products.

For an upgrader, the strategies are genuinely different animals, not variations on a theme:

• Buy first, bridge the gap. Certainty on the new property, exposure to the sale price and sale timing of the old one, peak debt and capitalised interest.
• Sell first, then buy. Certainty on the proceeds, exposure to the market moving while the client is out of it, and often a rental period.
• Simultaneous settlement. Lowest cost when it works, highest execution risk, and no slack if either side slips.
• Buy with a longer settlement or a deposit structure. Buys time without bridging cost, but depends on a vendor accepting the terms.
• Stay and renovate, or hold the existing property and buy as an investor. A different file entirely, and sometimes the right answer.

Choosing between those is a strategy recommendation. On ASIC’s framing, the reasoning behind it is part of what the duty covers, and it needs to be evidenced like any other part of the recommendation. RG 273.21 says ASIC expects evidence of compliance to “come predominantly from the broker’s records”.

Where upgrader files break: the sale assumption

Nearly every upgrader file rests on a number the client supplied and nobody tested: what the existing property will sell for, and how long it will take. In a rising market that assumption is forgiving. In a market where the Reserve Bank has just lifted the cash rate to 4.85 per cent and mortgage-holder sentiment has fallen to its lowest level since the Westpac–Melbourne Institute series began in 1994, it is not.

RG 273.40 is directly on point: “If it is reasonably apparent that the information about a consumer’s individual circumstances is incomplete or inaccurate, you should make further inquiries to obtain complete and accurate information.” RG 273.37 adds that where a consumer’s instructions seem inconsistent with their circumstances, you should make further inquiries. And RG 273.41 goes further again: “If critical information is not obtained when inquiring about a consumer’s circumstances, you should refrain from making a recommendation.”

For an upgrader, the expected sale price and the expected days on market are critical information. A client’s own estimate of what their house is worth, carried into a bridging calculation without testing, is the definition of information that is reasonably apparent as potentially inaccurate. The fix is not complicated: a documented range rather than a point estimate, an agent appraisal on file, a stated peak debt at the bottom of the range, and a recorded conversation about what happens if the property does not sell inside the bridging term.

Present more than one option — and record the one they didn’t take

RG 273.89 says that as a matter of good practice you should present a consumer with more than one option “unless there is a good reason not to”, assessed case by case. RG 273.20 says that where there are multiple options, they are to be presented in a manner consistent with the consumer’s best interests. RG 273.93 says it is important that consumers understand why the options presented were selected and why a particular option was recommended.

The record-keeping expectation is just as specific. RG 273.104(b) says you should clearly record the recommendations you presented, “including any alternatives that are not selected by the consumer — this should include the steps you took to educate the consumer about the options”.

On an upgrader file, that means the sell-first option your client rejected belongs on the file alongside the bridging facility they accepted, with a line on why they rejected it. That is not bureaucracy. If the sale runs long and the client later says nobody explained the alternative, the file note is the whole defence.

It is also worth remembering RG 273.18: brokers must exercise judgement, and “in some situations, this will include challenging the consumer’s perception of their best interests”. A client who has decided to buy first because they do not want to rent for three months has made a lifestyle choice, not a credit assessment. Testing that is part of the job, and recording that you tested it is part of the evidence.

What this means for your pipeline strategy

Concentration cuts both ways. A channel where upgraders are approaching half of flow is also a channel where upgrader expertise is the most commercially valuable thing you can be known for. The practical response is not to abandon the segment — it is to run it deliberately and to know what your second engine is before you need it.

Bailey’s expectation that refinancing could move from 15 to 16 per cent toward 20 to 25 per cent over the coming months is the obvious candidate, and it is a forecast rather than a fact. The Index’s own observation is that refinancing is being held down by serviceability constraints and lender retention activity, both of which are structural. A broker waiting passively for refinance volume to return may be waiting on something that does not arrive on schedule.

There is a geographic dimension too. AFG’s Q1 figures show lodgement value of $8.59 billion in NSW, $7.75 billion in Victoria, $4.91 billion in Queensland and $3.77 billion in Western Australia — with WA down 10.5 per cent on the quarter and 6.2 per cent year on year, its first year-on-year decline after 11 consecutive quarters of growth. Year-on-year declines in NSW, Victoria and Queensland ranged from about 13 to 14 per cent. If your book is concentrated in one state as well as one segment, that is two correlations stacked on top of each other.

What to watch next

The Q2 FY2027 Index, due in the new year, will be the first AFG read that fully covers the period after the November increase. The two numbers to watch are whether the upgrader share holds near its record or retraces toward the low 40s, and whether refinancing moves toward Bailey’s 20 to 25 per cent expectation. If upgrader share falls and refinancing does not pick up the slack, the channel’s volume problem becomes a good deal sharper than the current figures suggest.

The other thing to watch is lender appetite for bridging. If upgraders remain the dominant segment, bridging terms, end-debt assessment and peak-debt policy become competitive levers worth tracking across your panel.

The takeaway

A record 46 per cent upgrader share is not a sign of strength in the broker channel. It is a sign that the other segments have thinned while one held. For brokers, that has two consequences that need acting on separately: a commercial one, which is that your pipeline is probably less diversified than it was 12 months ago, and a compliance one, which is that the file type now dominating your desk is the one where the best interests duty reaches hardest — into the strategy, the assumptions behind it, and the options your client was shown and declined.

Neither is solved by writing more of the same files faster.

Sources

AFG Mortgage Index Q1 FY2027 (ASX announcement, 13 October 2026); AFG Mortgage Index Q1 FY2026; The Adviser, “Upgraders drive mortgage demand as first home buyers stall: AFG”, 11 November 2026; The Adviser, “Mortgage borrower sentiment sinks to all-time low”, 12 November 2026; Reserve Bank of Australia, Statement on Monetary Policy, November 2026; ASIC Regulatory Guide 273, Mortgage brokers: Best interests duty.