Mean Reversion: +3.0% Extra Growth Per Year
Suburbs that grew less than 44% over the past decade outperform by 3.0% per year over the next four years. This pattern held at every single sample date. That is extra growth, on top of whatever the market does.
This is one of several threshold indices in the Microburbs research programme.
What Is Mean Reversion?
Property markets revert to the mean. Suburbs that have grown quickly tend to slow down. Suburbs that have lagged tend to catch up.
This is one of the most consistent patterns in Australian property data. We tested it at 163 different time points across 14 years. The pattern held every single time. No other threshold in our research has a 100% consistency rate.
The threshold is straightforward. Measure how much a suburb's median house price grew over the past 10 years. If it grew less than 44%, it is in the top tier for future growth. If it grew more than 91.9%, it is in the bottom tier.
Low past growth is not a sign of a bad suburb. It is a sign of an undervalued suburb waiting to catch up. High past growth is not a sign of continued momentum. It is a sign that the easy gains have already been made.
Three Performance Zones
The model splits suburbs into three tiers based on their past 10-year median house price growth. Each tier shows a distinct forward growth pattern.
3.02% spread between top and bottom tiers
This is one of the strongest and most consistent signals in our research. It held at 100% of sample dates from 2008 to 2021.
Performance Over Time
The chart below tracks the 4-year annualised growth rate for low-past-growth suburbs (top tier, blue) and high-past-growth suburbs (bottom tier, red). The blue line sits above the red line at every single quarter.
Consistency Across 28 Sample Dates
We tested the signal at 28 different points in time between 2008 and 2021. The top tier outperformed at every single date. No exceptions.
| Sample Window | Extra Growth (4yr) |
|---|---|
| 2008 | |
| Mar 2008 → Mar 2012 | +2.63% |
| Sept 2008 → Sept 2012 | +2.53% |
| 2009 | |
| Mar 2009 → Mar 2013 | +2.43% |
| Sept 2009 → Sept 2013 | +2.02% |
| 2010 | |
| Mar 2010 → Mar 2014 | +2.32% |
| Sept 2010 → Sept 2014 | +2.41% |
| 2011 | |
| Mar 2011 → Mar 2015 | +2.77% |
| Sept 2011 → Sept 2015 | +3.63% |
| 2012 | |
| Mar 2012 → Mar 2016 | +4.07% |
| Sept 2012 → Sept 2016 | +4.63% |
| 2013 | |
| Mar 2013 → Mar 2017 | +5.49% |
| Sept 2013 → Sept 2017 | +6.61% |
| 2014 | |
| Mar 2014 → Mar 2018 | +6.38% |
| Sept 2014 → Sept 2018 | +5.67% |
| 2015 | |
| Mar 2015 → Mar 2019 | +4.84% |
| Sept 2015 → Sept 2019 | +3.35% |
| 2016 | |
| Mar 2016 → Mar 2020 | +1.62% |
| Sept 2016 → Sept 2020 | +0.56% |
| 2017 | |
| Mar 2017 → Mar 2021 | +1.25% |
| Sept 2017 → Sept 2021 | +1.29% |
| 2018 | |
| Mar 2018 → Mar 2022 | +1.95% |
| Sept 2018 → Sept 2022 | +2.61% |
| 2019 | |
| Mar 2019 → Mar 2023 | +2.34% |
| Sept 2019 → Sept 2023 | +1.57% |
| 2020 | |
| Mar 2020 → Mar 2024 | +2.26% |
| Sept 2020 → Sept 2024 | +2.93% |
| 2021 | |
| Mar 2021 → Mar 2025 | +4.34% |
| Sept 2021 → Sept 2025 | +6.32% |
Geographic Breakdown
The signal works across most Australian regions. The chart below shows the spread (low-past-growth suburbs minus high-past-growth suburbs) for each GCCSA region. Positive spread means the signal works as expected.
Full Regional Table
All growth rates are annualised over 4 years, measured 2008 to 2021.
| City | Spread | Sales Tested |
|---|---|---|
| Sydney | +6.40% | 63,808 |
| Hobart | +6.32% | 251 |
| Regional Tas. | +3.49% | 364 |
| Regional Qld | +3.30% | 119,654 |
| Regional SA | +2.78% | 26,495 |
| Regional WA | +2.77% | 41,480 |
| Brisbane | +2.63% | 48,015 |
| Regional NSW | +2.34% | 86,646 |
| Melbourne | +2.15% | 33,702 |
| Adelaide | +1.83% | 36,174 |
| Regional Vic. | +1.51% | 56,582 |
| ACT | -0.09% | 5,948 |
| Perth | -0.27% | 48,743 |
| Darwin | -0.97% | 3,999 |
| Regional NT | -1.80% | 1,358 |
Real-World Example: Denham Court vs Middleton Grange
These two suburbs sit 8 km apart in south-west Sydney. Both are new-build corridors with young families. The difference is past price history. Denham Court had underperformed for years. Middleton Grange had already boomed.
Denham Court, NSW 2565
Sydney
Past growth index: -0.258 (well below average)
Future growth vs median: +2.21% p.a.
Median hold CAGR: 12.9% across 9 property holds
Example property hold
20 Dogwood Crescent: bought February 2017 for $400,000, sold August 2021 for $872,000. That is 18.9% compound annual growth over 4.5 years. The suburb had been overlooked, then mean reversion kicked in.
Middleton Grange, NSW 2171
Sydney
Past growth index: +2.211 (far above average)
Future growth vs median: -2.05% p.a.
Example property hold
18 Reardon Avenue: bought April 2022 for $1,232,000, sold February 2024 for $1,235,000. That is 0.1% compound annual growth. Nearly two years of holding and the price went nowhere.
Is This Pattern Real?
We tested this rigorously. The pattern of +3.02% spread was confirmed by testing across 825,392 total sales over 14 years.
This is a real signal, not a crystal ball. Many factors drive property prices, from interest rates to local infrastructure to supply constraints. Across 14 years of data, this pattern holds with perfect consistency.
The signal worked at all 28 different time periods. It held in 12 of 15 geographic regions. The top tier outperformed the bottom tier in 100% of quarters. The t-statistic is 251.87 and the p-value is essentially zero. These results point to a genuine, repeatable pattern.
Want the Full Statistical Detail?
The Technical Whitepaper covers p-values, t-test methodology, and the full date-by-date and region-by-region breakdown.
Find Undervalued Suburbs Near You
Get mean reversion scores for every suburb in Australia. Identify suburbs where low past growth signals strong future performance.
Part of the Threshold Signals research programme