There’s more property data available to investors today than ever before, much of it free, and it can significantly improve investment decision making. Analysing how a property’s value has changed over time (i.e. its compound annual growth rate across multiple decades) and comparing it to comparable properties in close proximity, is an essential part of investment due diligence.
However, it’s equally important not to become too data driven. Over the past 5–10 years, there has been a surge in data-driven buyer’s agents, many of whom operate “borderlessly” across Australia. Investors should tread carefully. Data can inform good decisions, but overreliance on it, without understanding the nuances behind the numbers, can lead to very costly mistakes.
Proven past growth is one of three mandatory attributes
For a property to qualify as investment-grade, it must have three key attributes.
Firstly, it must possess genuine scarcity. That means there’s limited supply within an easily commutable distance – i.e., no vacant land nearby. And ideally, the property also has other scarce features such as period architecture, water views or a highly desirable street.
Secondly, more than half of the property’s total value must be in the underlying land. The higher the land-to-value ratio, the better.
Finally, it must have runs on the board. In other words, the property (and others like it close by) must have already demonstrated strong capital growth over several decades.
It’s this last attribute, a proven track record of capital growth, that I want to focus on today.
How do you build a reliable data set
Let’s say I’m interested in 13 Smith Street, a 2-bedroom Victorian house on 200 sqm of land that was cosmetically renovated within the past 20 years.
The first thing I would do is find its sales history of that property. I want to find every past sale and go back as far as possible. Using the XIRR formula in Excel (or an AI tool like Chat GPT), I can then calculate the property’s compound annual growth rate (CAGR) over different time periods. Generally, the longer the time frame, the more reliable the data.
Below is a table showing the sales history of a property located in an investment-grade street and suburb in Melbourne (picked at random). I have calculated the growth rate over several periods and benchmarked these results against the Melbourne’s median house price to provide context. For example, property price growth across Melbourne has been relatively weak over the past decade, so that context matters when interpreting the results. Overall, this property has performed well and has typically outperformed the median which should give us confidence that it could be a good asset.
The next step would be to undertake the same exercise for 11 and 15 Smith Street i.e., the properties on either side of the target property, assuming they are comparable in terms of land size and dwelling. And then make out way down the street doing the same thing. The more comparable properties you can find to do this analysis on, the more compelling the data set becomes.
However, this data can lead to a VERY expensive mistake
On the surface, this sounds like a straightforward approach. But in practice, the data can be misleading unless you understand the story behind the numbers. A purely data-driven strategy can easily lead to costly mistakes, such as buying the wrong property in the right location, or worse, the wrong property in the wrong location. Several factors can distort historical growth rates, and it’s important to recognise and adjust for these when assessing a property or location.
Timing can have a big impact
Using the example above, timing can have a significant impact on the calculated growth rate. For instance, the period between 2010 and 2022 produced only 4.5% p.a. growth, which is roughly half the rate achieved between 1986 and 2010, a timeframe that included several strong growth cycles.
The challenge with analysing property-specific data is that there are often only a few data points available. I deliberately used an example with four sales to illustrate the concept, but in many cases, there might be only two. This makes the timing of those sales crucial, especially if the period between them is less than 20 years, as it may only include one property cycle. If the first sale occurred just before a flat market cycle, the resulting growth rate will naturally appear lower than average because of a market wide phenomenon, not necessarily because the asset is inferior.
Benchmarking a property’s past growth against state medians and comparable properties helps broaden the dataset and provides useful context. However, it’s still essential to understand the market conditions that existed during each sale, whether the market was buoyant, soft, or recovering from an economic shock, as this context often explains much of the variation in results.
Unusual market activity
A location can sometimes be influenced by a temporary surge in demand from a single source of buyers. For example, a prominent buyers’ agent firm might start recommending a particular suburb, or a “hotspot” report might attract a wave of investors. This short-term demand can artificially inflate prices and distort apparent growth rates. Long-term locals usually recognise these anomalies, but they are far less obvious to outsiders who don’t have intimate knowledge of the area and are therefore overly reliant on data.
A smaller, yet still relevant, example occurred in October 2008, when the Rudd Government doubled the First Home Owner Grant from $7,000 to $14,000. Properties priced under roughly $600,000 experienced a noticeable, once-off spike in values as a result.
Fundamental flaws – no two properties are identical
Properties with fundamental flaws, such as poor privacy or security, road noise, bad orientation, proximity to commercial buildings, architecture that clashes with the surrounding streetscape, or an awkward floor plan, tend to underperform.
Two properties might look identical on paper in terms of measurable factors like land size or number of bedrooms, yet their capital growth can diverge significantly over time simply because one suffers from these flaws. Because no two properties are ever identical, it’s important to adjust for these qualitative considerations.
Capital improvements artificially overstate growth
A property’s physical condition plays a huge role in its capital growth performance. Has it been well-maintained, modestly updated, or fully renovated? Growth data does not distinguish between an unrenovated dwelling and a fully renovated one. If the owner has spent a lot of money improving a dwelling, that must be factored into any capital growth calculation. Otherwise, the growth rate will be artificially overstated because part of the increase in value reflects the renovation spend, not genuine market appreciation.
It’s not just about the renovation cost; it’s about buyer demand. In many markets, properties needing major work can be heavily discounted because few buyers want the hassle, even if the numbers look good on paper. And that sentiment changes over time. Since 2020, unrenovated homes have largely fallen out of favour as buyers worry about the difficulty of finding a reliable builder, not to mention today’s high construction costs. This sentiment can change so it’s important to keep this in mind when analysing sales data.
Location nuances
Aggregating sales data across even a small area can hide important differences. In some streets, one side may consistently outperform the other, perhaps due to better sunlight, superior views, backing onto parkland rather than commercial property, greater privacy or security, or inclusion within a preferred school zone. These nuances need to be considered when analysing the data. However, only someone with deep, long-term, on-the-ground knowledge may truly understand these subtleties.
Changes in zoning and heritage overlays
Planning changes can significantly influence a property’s appeal and value. For instance, the introduction of a heritage overlay might limit redevelopment potential, a restriction that some buyers view negatively, while owner-occupiers may see it as preserving character and scarcity.
Conversely, the relaxation of planning controls can increase the productive value of land and may trigger a one-off price spike as the market re-values the development potential.
Gentrification and significant changes in local amenities
A property’s past growth performance can be heavily influenced by changes within its suburb or surrounding area. For example, gentrification can transform a once-overlooked suburb into a highly desirable one, driving sustained increases in demand and, in turn, stronger capital growth. In such cases, historical growth rates prior to gentrification may understate future potential.
Conversely, newly established estates often experience strong early growth as essential amenities, such as schools, healthcare services, and supermarkets are added. However, once the area matures and supply stabilises, that growth typically flattens. In other words, the impressive past performance in these areas is unlikely to be repeated.
Enough datapoints to be statically reliable
If a suburb is tightly held, the available data may be extremely limited. Any calculated growth rate based on such a small sample is statistically weak and potentially misleading.
For growth analysis to be statistically meaningful, it should include at least 30 investigated transactions but ideally, more than 50. However, as mentioned earlier, not all sales are equally useful. Each transaction tells a different story, so it’s important to understand the context behind the numbers rather than relying solely on the headline data. Chances are that you are not going to find enough reliable sales data to draw a statistically meaningful conclusion, which is why you need to compliment that data with advice from a local area expert.
You need a local area expert AND the data
The key message is simple: you need a buyers’ agent with deep, long-term local knowledge, ideally someone who has been active in the same market for decades, to accurately interpret individual property growth data. A purely data-driven approach, without that local insight, is risky and can easily lead to costly mistakes.
We use this same approach when assessing clients’ existing investment properties. This allows us to form a view on their likely future returns and forecast what contribution each property will make towards funding retirement.
