What drives credit and house prices
Which assumptions about arrears and house prices hold up in Australia’s official data, what else matters, and how far a suburb’s price can be forecast.
General information, not legal or financial advice.
Which assumptions held
We tested eleven assumptions about what moves arrears and house prices against official data from the RBA, APRA, the ABS and Valuer-General Victoria. Four held, three held in part, three were not shown and one did not hold. An assumption held if its effect ran the expected way, with a t of 2 or more, and beat a naive guess on data the model had not seen. Of 40 factors nobody named, those that passed stricter tests are Found in the data.
| Assumption | Verdict | What the data shows |
|---|---|---|
| Inflation moves interest rates | Partly | The RBA responded from 2015 (t 4.6), but inflation forecast the cash rate worse than no change. |
| Interest rates move capacity to repay | Held | Mortgage rates move 0.98 points per point on the cash rate; arrears follow 4 quarters later (t 2.1). |
| Unemployment moves capacity to repay | Held | 0.10 points more loans non-performing per point of unemployment (t 13.2); it beat a no-change guess on later quarters. |
| Inflation moves capacity to repay, beyond its effect on rates | Not shown | No effect once rates and unemployment are known (t −1.5). |
| Interest rates and unemployment move house prices | Partly | Right direction; forecast first, no better than average growth (7.4% against 6.8% two years out). |
| The suburb's household income sets its prices | Held | Income alone explains 53% of price differences; socio-economic advantage carries the weight (t 16.1). |
| A new station, and convenience, lift prices | Partly | Distance to the CBD: −7.9% when it doubles. A new station: its suburbs grew 2.2% less than matched suburbs (95% interval −4.4% to +0.4%) two years after it opened. |
| Land sets prices | Held | Councils’ median house block: +19.1% when it doubles (t 7.3). |
| A new centre, business hub or precinct lifts prices | Not shown | Suburbs near a new centre, hub or precinct grew 5.0% less than matched suburbs (95% interval −8.5% to −1.8%) three years after it was announced; the newer projects did not repeat it. |
| Natural resources set prices | Not shown | No open measure; mining jobs add nothing (t 1.8). The coast adds 21% (t 8.2). |
| The same factors say which suburbs will grow | Did not hold | Beat average growth by 4% on later years; four features changed sign. |
| Found in the data | ||
| Repayments relative to income: the price level | Found | +19.6% per standard deviation (95% interval +12.5% to +27.1%) |
| Residents aged 25 to 39: the price level | Found | −11.5% per standard deviation (95% interval −13.2% to −9.8%) |
| Density: the price level | Found | +11.0% per standard deviation (95% interval +8.0% to +14.1%) |
| Households renting: growth | Found | −0.60 points a year per standard deviation, in both periods |
| New loans at 80% LVR or more: arrears ahead | Found | Leads by 7 quarters; 24% closer than “no change” |
| Dwelling prices: arrears ahead | Found | Leads by 4 quarters; 11% closer than “no change” |
| A region’s insolvency rate: next year’s | Found | 38% closer than assuming the state’s rate |
What matters most
- Point for point, unemployment is the strongest link to mortgage arrears. Each point has gone with about 0.10 points more housing loans non-performing, against 1.01% now (APRA; ABS).
- Rate rises reach borrowers almost in full, and arrears follow about a year later. A $600,000 loan’s repayment went from $2,530 a month in December 2021 to $3,675 in July 2026 (RBA table F6).
- Arrears concentrate where equity is thin. Loans at 95% LVR or more have been non-performing 11.7 times as often as loans under 60% (APRA).
- A suburb’s price level can be explained; its growth cannot. Socio-economic advantage, distance to the CBD, the coast and land explain 80% of the differences between suburbs’ medians on councils the model had not seen; with three of the factors below, 89% (Valuer-General Victoria; ABS SEIFA).
The ten factors that track prices most
Each factor’s correlation with suburbs’ 2025 median house prices, one at a time, strongest first (ABS Census; Valuer-General Victoria). Factors overlap, so the last column says whether each still matters when tested with the others. Rent and repayments are left out: they are partly the price itself.
| Rank | Factor | Correlation | Goes with | Still matters with the others |
|---|---|---|---|---|
| 1 | Socio-economic advantage (SEIFA IRSAD) | +0.83 | Higher prices | Yes |
| 2 | Share of adults with a bachelor degree or higher | +0.82 | Higher prices | No |
| 3 | Distance to the CBD | −0.77 | Lower prices | Yes |
| 4 | Median household income | +0.74 | Higher prices | No |
| 5 | Outside Greater Melbourne | −0.65 | Lower prices | Yes |
| 6 | Residents per square kilometre | +0.64 | Higher prices | Yes |
| 7 | Share of homes that are separate houses | −0.56 | Lower prices | No |
| 8 | Council median shop price per m² (proxy) | +0.55 | Higher prices | No |
| 9 | Share of land prone to bushfire | −0.54 | Lower prices | Yes |
| 10 | Distance to the nearest railway station | −0.53 | Lower prices | Yes |
What else matters
Tested on councils and years the models had not seen, with one correction for testing 40 factors, these carry weight (ABS Census; Valuer-General Victoria; method). Each is associated with prices; none is shown to cause them.
- Repayments high relative to income. Where repayments take more of local income, prices are higher (+20% per standard deviation). Partly that is the price itself; it also marks buyers who stretched further, with thinner buffers.
- More residents aged 25 to 39. Such suburbs are cheaper than location and income suggest (−11% per standard deviation); valuing them by their neighbours can overstate them.
- Density. Denser suburbs are dearer than distance and income explain (+11% per standard deviation).
- More households renting. Such suburbs grew a little more slowly in 2016–2021 and 2021–2025 (−0.6 points a year per standard deviation); small, 0.5% closer than the starting price alone.
Early warning
What moved first in APRA’s 30 quarters of housing data, and AFSA’s regional insolvencies (APRA; AFSA). One cycle: a guide, not a rule. Less high-LVR lending marks tighter credit, not safer loans.
| What to watch | What followed | For | Checks passed |
|---|---|---|---|
| More loans 30 to 89 days behind | Arrears up 2 quarters later | Home loans | 1 of 3 |
| Fewer job vacancies | Arrears up 1 quarter later | Home loans | 1 of 3 |
| Fewer new loans at 80% LVR or more | Arrears up 7 quarters later | Home loans | 3 of 3 |
| Dwelling prices falling | Arrears up 4 quarters later | Home loans | 3 of 3 |
| A region’s insolvency rate, against the state’s | 88% of the gap stays next year | Personal and small loans | 1 of 1 |
| A jump in a region’s insolvency rate | 38% of it reverses next year | Personal and small loans | 0 of 1 |
Six example suburbs
Illustrative only. Not a valuation, not a prediction to rely on, and not advice.
Each suburb’s 2025 median moves with its market, at a standard variable rate of 8.77% and the RBA’s August 2026 forecast. Ranges run from the 5th to the 95th percentile of past misses; on years left out, they held 79% of two-year outcomes and 74% of three-year ones (Valuer-General Victoria). Adding the growth factor that passed did not improve them.
| Suburb | 2025 median | 2027 range | 2028 range | Past two-year misses, inside the range |
|---|---|---|---|---|
| RichmondInner | $1,397,500 | $1.21m to $1.94m | $1.21m to $1.96m | 13%; 9 of 10 |
| ReservoirMiddle-ring | $950,000 | $853k to $1.30m | $856k to $1.36m | 15%; 7 of 10 |
| CraigieburnOuter growth | $715,000 | $658k to $1.02m | $676k to $1.07m | 11%; 9 of 10 |
| South MorangNear new rail | $800,000 | $736k to $1.14m | $757k to $1.20m | 12%; 9 of 10 |
| MilduraRegional | $536,000 | $481k to $786k | $487k to $866k | 8%; 10 of 10 |
| TorquayCoastal | $1,202,500 | $1.08m to $1.76m | $1.09m to $1.94m | 17%; 7 of 10 |
How wrong it has been. Since 2005 the method missed Melbourne by 7.4% two years out, no better than average growth (6.8%); early 2026 sales are below the bottom of its 2026 range (Valuer-General Victoria). The ranges are the result, not the midpoints.
New stations and centres add little the data can see. With the event study’s small and uncertain effect, Glen Waverley’s 2027 range (Suburban Rail Loop East) moves from $1.55m–$2.36m to $1.57m–$2.41m, and Point Cook’s (East Werribee Employment Precinct) from $759k–$1.17m to $776k–$1.20m (method; Valuer-General Victoria).
What it means for a lender
- Serviceability buffers cover rates, not job loss. Loans written at the 2021 low were tested at 5.6%; borrowers now pay 6.2% on average (RBA table F6).
- Collateral risk differs by kind of suburb. Over two years, the middle 90% of suburbs strayed from their market by −14% to +18% in inner Melbourne and −15% to +24% in regional Victoria (Valuer-General Victoria).
How we tested this
Every model, its test and what the data cannot show: the method.
Sources
- RBA statistical table Indicator Lending Rates (F5): standard variable housing rate, owner-occupier. Reserve Bank of Australia, CC BY 4.0, retrieved 4 October 2026. Data: f5-data.csv.
- Quarterly Authorised Deposit-taking Institution Property Exposures statistics, June 2026. Australian Prudential Regulation Authority, CC BY 4.0, retrieved 4 October 2026. Data: Quarterly authorised deposit-taking institution property exposures statistics June 2026.xlsx.
- Census of Population and Housing 2021, General Community Profile DataPack, Suburbs and Localities, Victoria. Australian Bureau of Statistics, CC BY 4.0, retrieved 4 October 2026. Data: 2021_GCP_SAL_for_VIC_short-header.zip.
- Victorian Property Sales Report — Median House by Suburb Time Series, 2015–2025. Valuer-General Victoria (Department of Transport and Planning), CC BY 4.0, retrieved 4 October 2026. Data: houses-by-suburb-2015-2025.xlsx, from the Internet Archive’s copy.
- RBA statistical table Labour Force (H5): unemployment rate. Australian Bureau of Statistics, via the Reserve Bank of Australia, CC BY 4.0, retrieved 4 October 2026. Data: h5-data.csv.
- RBA statistical table Housing Lending Rates (F6): owner-occupier variable rates, new and outstanding. Reserve Bank of Australia, from APRA data, CC BY 4.0, retrieved 4 October 2026. Data: f6-data.csv.
- Socio-Economic Indexes for Areas (SEIFA) 2021, Suburbs and Localities. Australian Bureau of Statistics, CC BY 4.0, retrieved 4 October 2026. Data: Suburbs and Localities, Indexes, SEIFA 2021.xlsx.
- Quarterly Personal Insolvency Statistics — regional time series by Statistical Area Level 3, from September 2007. Australian Financial Security Authority, CC BY 2.5 AU, retrieved 4 October 2026. Data: regional_quarterly_time_series.csv.
- Statement on Monetary Policy, August 2026 — Outlook, Table 3.1 Detailed Forecast Table. Reserve Bank of Australia, CC BY 4.0, retrieved 4 October 2026.
- Victorian Property Sales Report — Time Series, Year Summary 2025 (by municipality, 1995–2026). Valuer-General Victoria (Department of Transport and Planning), CC BY 4.0, retrieved 4 October 2026. Data: year-summary-2025.xlsx, from the Internet Archive’s copy.
- Every source used, with its licence and the day it was retrieved: the method’s sources.