32% was the Optimistic Number

Ray White modelled that Brisbane property prices would need to fall 32% to bring investors back to the market under the quarantined negative gearing rules — if rents hold flat. I wrote about that number a week ago and it is the most-read thing I have published.

I want to revisit it, because the assumption sitting underneath it has just changed. Because there is a third outcome that number does not contemplate at all.

And because a 32% fall in residential values is not a property story. It is a credit and legislative story, and it reaches a long way past the people who own investment properties.

The arithmetic nobody spells out

An investor buys on yield. Yield is rent divided by price.

Quarantining negative gearing removed a tax shield that let investors accept a lower cash yield, because the shortfall was subsidised elsewhere in their return. Take the shield away and the yield has to do the work on its own. That is the entire mechanism behind the 32%.

There are only two ways to move a yield. The numerator goes up, or the denominator comes down. Rents rise, or prices fall.

That is the whole equation. Everything else is commentary.

Path one is open, but it is bounded

Rents are moving, and faster than the commentary suggests. Brisbane house rents ran 2.9% in the June quarter to a record $700 a week — around four times the pace of the prior quarter, and the fourth consecutive quarterly rise. Vacancy sits at 0.6%, equal lowest June on record.

So the question is not whether rents can rise. It is whether they can rise far enough.

If a 32% price fall restores the yield, then closing the same gap through rent alone takes the reciprocal. Roughly 47%, or about $1,030 a week on today's median.

Now put an income against it. Australian renters are already at a record 33.1% of gross household income, above the 30% stress threshold and up from 26.2% in 2020. For rent to reach $1,030 while renters hold their current share, household income has to rise 47% as well. At 3.5% wage growth, that is a shade over eleven years.

Anything faster means renters absorbing a larger share of income than any cohort has sustained at a median level.

And the cheaper segment has already stalled. Brisbane unit rents sat flat at $660, ending six consecutive quarters of growth, with annual growth at a 15-month low. Same city, same 0.6% vacancy, opposite result by dwelling type. That is not a supply story, because supply is identically tight for both. Units are where renters go when houses get away from them, and the cheaper option reaching its limit first is what an affordability ceiling looks like on arrival.

One quarter is not a trend. But it is the print I would want repeated before assuming rents carry this.

Does the required fall get bigger than 32%?

Ray White's modelling was done in a particular rate environment. That environment moved last week.

July CPI came in at 3.5% annual, down from 3.8%. On the headline that reads like progress. The number the Reserve Bank actually watches, trimmed mean, held flat at 3.6% — outside the 2–3% target band, and giving the Bank nothing to work with.

NAB has now joined CBA and ANZ in forecasting hikes, flagging a rise at the 29 September meeting with a second possible in November, which would put the cash rate at 4.85%. That is the highest setting since the Global Financial Crisis. Three increases this year have already added roughly $272 a month to a $600,000 mortgage, at about $91 for every 25 basis points.

Here is why that matters for the 32%.

A higher cash rate raises an investor's cost of capital. A higher cost of capital raises the yield that investor requires before a deal clears. If the required yield rises while rents are bounded by wages, more of the burden falls back on price.

I am not putting a new percentage on it. That is Ray White's model, not mine, and I have not run it. But the 32% was struck on a set of assumptions, and the rate assumption has moved.

The inflation problem is not easing

Fuel is the part people feel and forecasters keep underestimating. The national average unleaded price for August was 209.5c a litre, up 13.6 cents in thirty days. Diesel is running near 252c. Diesel is the one that matters structurally, because it moves freight, and freight moves everything else in the basket.

Housing was the largest single contributor to annual inflation, rising 5%.

An inflation problem whose two biggest contributors are fuel and housing is not one interest rates fix quickly. It is one they fix slowly and expensively.

The outcome the 32% does not contemplate

Everything above assumes the market clears at some price. There is a third possibility, and the most experienced auctioneer in the country has been describing it for a month.

Tom Panos headlined his own market wrap in early August with a single phrase: the market is in gridlock.

He had one auction scheduled that Saturday. He normally conducts ten, twelve, sometimes fifteen. That one auction drew zero registered bidders, the second Saturday in a row he had taken no registrations at all. Forty years in real estate, thirty of them auctioning, and his read was that this is the lowest level of auction activity he has personally experienced — "I think this is the lowest since 1991."

Worth knowing what 1991 actually looked like, because it was not a crash. I bought my first home that year. Residential prices fell single digits nationally, and Brisbane rose 6.8%. What did happen was that the market stopped. Melbourne took until 1996 to get back to its 1989 nominal price. Seven years for a round trip to nowhere, with inflation eating the difference. That is what a freeze looks like from the far side, and it is a very different thing to model than a correction.

His description of who is missing is the part worth reading twice. "Investors are waiting. First-home buyers are waiting. Vendors are waiting." Developers too. Not one cohort withdrawing. All of them, at once.

And on the mechanism: "Markets don't collapse overnight. They freeze and they've frozen."

He also made the point that this happened in a week when the inflation print came in better than expected, and confidence did not move regardless. That is the detail that should worry people. If good news does not restart a market, the problem is not the news.

Two caveats, because they matter. Panos works in Sydney, where values have already fallen further than Brisbane's, so his floor is not a national one. And Brisbane is running its own version rather than the same one — clearance dropped to 30.5% in late July, on track to finalise below 40% for a ninth consecutive week.

But the mechanism travels, and here is why it is not just commentary.

The negative gearing changes are grandfathered. Existing investors are not forced sellers. If a vendor's expectation sits at a number a buyer will not fund, and nothing compels that vendor to transact, the market does not reprice.

It stops trading.

Why a freeze is worse than a correction

Panos made one more observation that most of the coverage skipped past. When transactions stop, the pain spreads outward — to agents, brokers, conveyancers, tradespeople, and eventually to renters.

That is the sentence that turns this from market commentary into an agency problem.

A 32% correction is a market that still works. Prices reset, buyers return at the new level, investors exiting create listings, and an agency writes business on both sides of the move. Painful, but transactional.

A freeze lists nothing and sells nothing. No repricing, no exits, no stock. Gross commission does not fall gradually — it stops arriving, and the listings that would normally replace it never come to market.

If you are modelling a downturn, those are two completely different businesses.

This is a credit and legislative event, not a property event

Here is where it stops being a real estate story.

A little under half of small business loans in Australia — those where the lender's exposure sits under $1.5 million — are secured by residential property. Not commercial property. Not equipment. Not the business itself. The owner's house, or their investment property.

And it is not a marginal preference. RBA research found that new small business loans secured against residential property are on average four and a half times larger than loans secured any other way. Residential security does not just make borrowing cheaper. It determines how much a business can borrow at all.

The Reserve Bank has been explicit that this cuts both ways. In its March Financial Stability Review it noted that information on small business leverage is limited precisely because so many owners borrow against their own residential property — and that continued housing price growth had likely been supporting those balance sheets.

Read that in reverse and you have the part nobody is discussing.

If residential values fall materially, the loss is not confined to investors and it is not confined to agencies. It shrinks the collateral base sitting underneath a large share of the small business credit in this country. Facilities that were written at a valuation stop supporting the same limit. Redraw capacity that funded working capital compresses. Refinancing that assumed a revaluation no longer settles. From there it moves to sub-prime lending and the cost of capital increases substantially. The business is now under pressure.

None of that shows up on a P&L. It shows up the day a business needs its facility and finds the number has moved.

And the price of that credit was already climbing before any of this. Small business variable-rate lending secured by residential property is sitting around 7.00% — before the hikes now being forecast for September and November.

So the answer to "who does a 32% correction affect" is not investors. It is every business in the country that borrowed against a house.

Which puts an agency principal on both sides of it at once. Their revenue comes from residential transactions, and their own facility is secured against residential property. Very few businesses carry that exposure twice.

What is actually carrying the agency

In the freeze scenario, the rent roll is the business.

It is the only part of an agency producing recurring income, and the only part carrying a value someone else will pay for. Sales walks out the door with the people who built it.

And the rent roll is squeezed from both ends.

Management fee income is a percentage of rent collected. If rents are bounded by what households can pay, so is your rent roll's income growth — years of getting revenue growth for free, without adding a single management, is not a permanent arrangement.

Rent rolls are valued and financed on a multiple of that income. If you borrowed to acquire one, the debt service on that facility reprices upward with the cash rate while the income servicing it runs into a wage-growth constraint.

Rising cost. Bounded income. An asset valued on a multiple of that bounded income.

Three sides closing at once, and they arrive before the volume decline does.

Most have probably already modelled the sales side of a downturn. Very few have run their rent roll facility at a 4.85% cash rate against management fee income growing at the rate wages allow.

And almost nobody outside real estate has asked the simpler version of the same question: if the house securing my business facility is worth materially less, what does my lender do at the next review?

Those are the numbers worth having before the September meeting, not after it.

John King is a Fellow of the Institute of Public Accountants and an Authorised Credit Representative. Peer Group Advisory works with boutique real estate agency principals on forward cash flow, structuring and acquisition funding.

No principal is going to work through their rent roll debt position in a comment section. If you are quietly running that number right now, send me a DM.

Peer Group Advisory