Investors have access to reams of data about Australia’s property markets. But what does it all mean and how can investors use it to maximise their financial success? By Matthew Liddy
There’s a plethora of statistics available about Australia’s property scene. Median house prices, median advertised rents, indicative gross yields, rental vacancy rates, days on market, auction clearance rates, and the list goes on.
But do you understand what all these numbers mean and how they should influence your investing?
Many investors find the lingo of property data too confusing; others are puzzled by the statistics’ sometimes-contradictory nature.
The first rule about pretty much all the property statistics available is to use them as a guidebook, not a rulebook. None of the data is perfect. A median price doesn’t represent every property in a suburb. Auction clearance rates might represent the health of a particular segment of the market, but they don’t tell you why some properties are selling when others aren’t.
That’s not to say that studying the data isn’t useful for property investors. It’s just that investors need to go one step further before making a purchase by doing a little “shoe leather” research, the kind that involves getting out and about to investigate a local market.
Buying on the basis of statistics alone can lead investors down the wrong path, Wakelin Property Advisory director Monique Wakelin says.
“I know people who’ve made horrible mistakes in the property market,” she says, “and all they do is sit behind a computer and look at median values and chart those, and then they go out, sight unseen almost, to a property and buy it based on what they reckon the median value for that suburb is… That’s just a classic mistake.”
Wakelin warns investors not to get too caught up in what any individual piece of data might mean about a suburb.
“You don’t want to isolate one factor at the exclusion of everything else that makes a good investment,” she says.
Demographic data, for instance, might be “one small part of the equation” but Wakelin says she’d never buy a property on the basis of that particular kind of statistic alone.
Property Planning Australia director Mark Armstrong tracks property data and knows that if used correctly it offers valuable insights into the market. But he too warns investors that pure statistics can only go so far.
“A lot of this stuff is broad-brush stuff. It’s a very broad analysis, and the only way to really know what to buy is actually get into the market and do that groundwork,” he says.
Margaret Lomas, property author and chair of the Property Investment Professionals of Australia, is another analyst who conducts detailed analyses of multiple property markets around the country. But she too notes property investment is about much more than data.
She says there are “must-ask questions” for every purchase and data is only involved in a couple of them.
In other words, property data is a powerful tool but it should be wielded with caution – and manipulated only by knowledgeable, educated users.
To help in that education process, API has created a comprehensive guide to what property statistics are available, where to find them and what they mean.
There are essentially two types of data available about the property market. The first is data designed to provide an insight into the health of an entire property market, be it at national, state or suburb level. The second is data designed for use when making individual investment decisions.
The data presented in this guide is roughly presented in the order of where it sits on the continuum between these two types of figures, with the broader market statistics presented first.
Median house and unit prices are probably the statistical measure property investors are most familiar with; they’re also easily misunderstood, misinterpreted and misused.
The median price is a way of measuring typical sale prices for property in a given location. It's the middle value of all sales over a given timeframe.
If, for example, 11 houses sold in the suburb of 'Exampletown' then if you ranked the sale prices from lowest to highest the sixth sale would provide the median price.
Medians aren’t a perfect reflection of property prices, but they’re the best measure available.
API readers are probably most familiar with the median price data presented in the Databank section at the back of each month's issue. The median prices presented there cover a 12-month period. This month API presents data for the “12 months to 31 December 2009”, meaning the medians are a representation of all the sales during the period between January 1, 2009 and December 31, 2009.
If there are fewer than 10 sales in a suburb during the given 12-month period then that suburb’s median price is considered “statistically not reliable”, and it’s excluded from the results.
API’s data is sourced from Australian Property Monitors (APM), which gets access to sales data from the valuer-general’s office in each state and territory, as well as from real estate agents.
Take the data published this month for houses in Alawa in the Northern Territory as an example (Alawa is the first suburb listed in the NT houses data section, on page 129).
The data shows APM has tracked 39 house sales in Alawa over the 12 months to December 31, 2009. The median price recorded is $470,000, meaning that if you listed the 39 sales from the lowest price to the highest, the middle price (the 20th sale on that list) is $470,000.
Monique Wakelin explains the median price recorded for a suburb can be skewed by changes in the type of property that’s been sold.
“For a particular time period, like for example while we had the boosted First Home Owner Grant, there were a lot of transactions in the lower price brackets and what that did was skew the median values,” she explains.
“Conversely, when conditions get very strong and these grants sunset, there’s a bias of transactions in the middle and upper range, and so that skews the median value.”
Rismark International managing director Christopher Joye says this is just one of a number of biases that can affect median price data.
For example, if people build and transact bigger or smaller homes over time, then the median price may rise or fall, giving the impression that prices have climbed or fallen, when in fact they really haven’t.
Renovations also cause problems for median prices, Joye notes, because if a lot of homes are renovated in a given suburb, that can artificially lift the median price, suggesting there has been capital growth, when in fact prices are higher because the houses being sold are better than they were previously.
Beyond the level of suburb-level median prices, a number of independent providers examine house and unit prices at a capital city level. These data houses include APM, RP Data, Residex and the Australian Bureau of Statistics.
Each of these providers has different methodologies, each with their own strengths and weaknesses. However, generally speaking, they show price trends over the longer term in a similar light.
It’s worth noting that median prices are always a historical measure; it takes time to collate all the prices being paid for property and put together a dataset that provides a reliable median price. Thus, most median price measures are at least three months old before they’re available for publication.
Despite the biases and issues involved, Joye says medians remain useful.
“The median is useful if you want to simply know what the middle sales observation in, say, Melbourne was over the past, say, quarter,” he says.
“This gives you a quick and easy-to-understand… guide for the price of the homes being purchased in the market…
“Median prices are also useful when seeking to address research questions that are targeted at identifying a ‘representative’ price at any particular point in time.”
Lawless agrees that median prices have their place, even though they might bounce around a little bit.
“They do provide a pretty good level of relativity in the market, in terms of what prices are being paid in a particular market in a particular timeframe,” he says.
Median price growth
The next step up the data tree from a raw median price is to look at how the median price in a particular location changes over time.
There are two sets of figures in API’s Databank that fit this description. First up is median 12-month growth, represented as a percentage.
Let’s return to the example cited earlier of Alawa in the NT. Its median 12-month growth comes in at 11.6 per cent, which means its median price in the 12-month period ending on December 31, 2009 is 11.6 per cent higher than for the corresponding period ending on December 31, 2008.
The next median price growth figure in API Databank is annual growth over 10 years, again shown as a percentage.
For Alawa, this comes in at 11.9 per cent, which means that over the past 10 years Alawa’s median house price has grown at an average rate of 11.9 per cent per year.
If all those ‘medians’ and ‘averages’ leave you confused, think of it like this: APM has taken the median price growth for each year over the past decade, added them up and then divided the sum by 10. This provides an indication of the rate of growth over a longer timeframe.
As you’d expect, all the same biases and limitations that affect the collection of straight median house prices also affect data for changes in those median prices over time.
Here’s how Tim Lawless explains it: “Using medians to determine capital growth can be a little bit misleading, particularly if you’re looking at areas that have a differentiated stock offering.
“What I mean by that is if you have a suburb that has a lot of waterfront housing or elevated housing that would have a significant premium attached to it, and on the other side you have houses that don’t have those premium attributes, if you find that a lot of waterfront stock is selling it’ll bring the median price up – and vice versa.
“So in areas that don’t have that homogenous housing stock, you’ll find that the median’s going to be a lot more volatile and provide less indications of how prices have actually moved over time.”
Lawless adds, “Before making an investment decision you really need to be drilling down a little bit further (than median prices) and actually looking at comparable sales, which is probably the real key to understanding what a particular property should be bought for.”
Wakelin says it’s best to examine changes in median prices only in annual periods (as presented in API Databank), rather than quarter-to-quarter, because the quarterly data can bounce around too much and doesn’t smooth out the anomalies to a sufficient degree.
“I use median values, and I use annual data, to give me a broad, long-term trend – so overall the market is either going up, down or sideways,” Wakelin notes. “Then I use the daily press to say, I know that one-bedroom apartments in Armadale are selling between this price and that price.”
Armstrong also uses median prices for broader research.
“Median data is really good to work out where we are in the broader market cycle, and looking at it over the long term,” he says.
“But we don’t use median data at all when we’re looking at individual asset selection. I think if you’re relying on median data for asset selection you’re really not getting deep enough into the analysis that needs to be done.”
It’s important to understand that all property in a suburb isn’t the same, and as a result median prices can’t reflect everything that’s happening in the market, Wakelin adds.
“That’s why you never get two one-bedroom units selling for the exact same price – because they’re not the same.”
Changes in median prices are by their very nature historical data and for this reason, Margaret Lomas says the data has a limited application.
“It can only tell me what has happened, rather than what will happen,” she says.
“Also, it doesn’t usually reveal why something happened. For example, strong growth data may be due to a number of things, such as a short-term infrastructure project, and this impetus may not be sustainable into the future.”
Investors need to pair median price movements in a suburb with other data and on-the-ground research before making an investment decision, Lawless says.
Different investors will use historical median price movements in different ways. While some might view strong median price growth over recent years as an indicator of a growing area,
others actually seek out areas that have had poor performances, as they believe that sluggish growth might indicate the potential for stronger catch-up growth in years to come.
Experienced investors find that median price statistics are very useful, but only as guides. You can’t take a figure for median price growth in a suburb and assume that your property’s value will have climbed by the exact same proportion, for instance.
Thus, investors shouldn’t rely solely on past median price growth when deciding where to invest, as past results don’t guarantee an area will continue to experience capital growth. Median price growth is, rather, best used as a starting point for research on where to invest and what to buy.
It’s never the replacement for solid due diligence and on-the-ground research.
Median rental values
Median rents are designed to provide an indication of rentals for properties in a particular location.
They work in a very similar manner to median prices, except they’re covering a different data set – rentals rather than sales. As you’d expect, they suffer from many of the same biases as median sale prices.
The median rent in a suburb won’t tell you what a top-end property will rent for, nor a low-end property, though it should give a hint at what a typical property might fetch in the rental market.
Monique Wakelin says median rents are much the same as median prices, in that investors should use them as a long-term guide only.
API’s Market Watch section, published each month, provides median advertised weekly rental rates for thousands of suburbs across the nation, provided by APM. The median rent in a suburb is the mid-point of all advertised rents in the given suburb over the stated 12-month timeframe.
Returning to our earlier example of Alawa in the NT, this month’s data shows a median rental value of $475.
APM’s Clinton McNabb explains that all rental information is based on advertised rents the company collects online.
“Only the most recent advertised rent amount is used to reflect as close as the rental amount as possible,” he says.
“All APM data are validated… and de-duplicated as we often collect the same data from various sources. All agency-reported data are also washed against the valuer-general’s results, which arrive at least eight weeks after.”
Median rental yields
The rental yield on a property is the return an investor can expect to receive in rent, expressed as a percentage of the purchase price of the property.
The rental yield on an individual property is calculated by multiplying the weekly rent by 52 (for weeks per year), dividing that figure by the purchase price of the property, and then multiplying the result by 100 to create a percentage. The resulting yield is a gross figure, meaning it doesn’t take into account tax; it also doesn’t account for any of the additional costs of holding a property.
Median rental yield statistics put together two of the key pieces of data discussed above – median prices and median rents – to create an indicative rental return figure for properties in a given location.
API publishes median rental yields for thousands of Australian suburbs in its Market Watch section at the back of each edition of the magazine.
Returning to our example suburb, Alawa in the NT, if you take the median advertised rental figure of $475 and the median house price of $470,000, and follow the rental yield calculation outlined above, you arrive at an indicative gross rental yield of 5.3 per cent. This would be considered a relatively strong rental return compared to many parts of the country.
Given that calculating median rental yields uses two sets of data (median prices and median rents) that we know have some in-built biases, it’s clear that the yields presented are best treated as an indicator of yields in the suburb. Investors shouldn’t assume that all properties in the suburb will provide similar returns, because it’s simply not realistic.
Still, median rental yields can prove useful as a guide to suburbs that could provide a potentially strong yield.
Days on market
Average days on market statistics are a measure of how long it takes to sell property in a given location once it’s been advertised.
API publishes days on market data at a suburb level in the Market Watch section at the back of each issue of the magazine, as well as listing days on market statistics for eight major Australian cities in its Databank section.
Looking at our example suburb of Alawa, we see that the average time it takes to sell a property in that NT suburb is 89 days, which most investors might regard as a lengthy period of time.
API’s days on market data is provided by APM, which calculates the average days on market for each of those locations by taking the difference between the date of the initial advertisement for the sale of the property and the date of the exchange of contracts on that property.
The number of days property sits on the market before sale will give investors a measure of supply and demand in those markets. It may provide some insight into what times of year property sells faster, and the relative strength of a market at any point in time.
Lawless rates how long properties take to sell as one of a number of key “leading indicators” for property, alongside the level of vendor discounting, the amount of stock on the market and the number of sales in trend terms.
These measures, Lawless says, are useful for analysing the health of a property market.
He rates them as “leading indicators” because compared to median prices, which are a measure of how a market stood at some point in the past, these statistics can provide some insight into where a market is headed.
Although it’s regarded here as a broader measure of analysing a property market, the number of days a particular property has been for sale can also provide an indication to a buyer of the level of demand for that property.
If a property has been on the market for a long time, it may suggest the vendor will be getting desperate, providing the opportunity for tough negotiation on price and allowing the buyer to get a better deal. However, Armstrong warns this isn’t always the case.
“You could also be looking at a really compromised and dud asset, which could be equally as true,” he notes.
Stock on market
Stock on market, as the name suggests, provides a measure of how many properties in a given location are currently being advertised for sale at any one time.
The number of listings can provide an indication of the health of a property market, as a sustained rise in the amount of stock on the market may suggest an oversupply situation is developing, with a lot of properties advertised for sale but few willing to buy. Conversely, a drop in advertised listings could suggest the development of a tighter market.
Seasonal factors will also come into play. For instance, listings might be few and far between around Christmas, but much more common in spring. This isn’t necessarily a sign of a tight or oversupplied market, but more just a representation of the normal state of affairs given when people choose to sell their properties.
Investors should therefore keep seasonal factors in mind when analysing stock on market figures. Comparing like-for-like figures will help minimise the seasonal issue (i.e. comparing January 2010 figures with January 2009 figures, rather than comparing January 2010 figures with December 2009 data).
API publishes stock on market figures for postcodes around the country in its Databank section at the back of the magazine (this month’s stock on market data starts on page 112). API’s data, sourced from SQM Research, provides annual comparison points, which help to minimise the seasonal issues.
Returning to our example suburb of Alawa in the NT, it comes within the postcode 0810. For this example, we’ll use the stock on market data published on page X, which shows that the postcode 0810 had X houses and Y units available for sale in (insert date), compared to X houses and Y units available for sale in (insert date).
This means total stock on market was X per cent lower in (insert timeframe) than it was in (insert timeframe). Looking at the numbers themselves, you could deduce that little has changed in terms of how many properties are listed for sale in and around Alawa in (insert timeframe) compared to (insert timeframe).
Tim Lawless ranks stock on the market as another of his “leading indicators” for property in a given location.
He also takes his analysis a step further by comparing the stock on the market to the total number of houses or units in a particular suburb.
Investors can do the same by finding the total number of established dwellings in a given postcode (this is available from the Australian Bureau of Statistics’ (ABS) most recent Census data, compiled in 2006).
To calculate the ratio of available stock to the total dwelling count, divide the stock on market figure by the total established stock figure and then multiply by 100 to obtain a percentage.
“If you see there’s, say, 80 per cent of stock on the market, that’ll set a few alarm bells off,” Lawless says.
For our example of Alawa and postcode 0810, the ABS shows there are 11,299 total private dwellings in the postcode. Thus, the ratio of stock on market to total established stock in (insert date) is calculated in the following manner: X (houses on the market) plus Y (units on the market) equals Z (total dwellings on the market), divided by 11,299 (total private dwellings), multiplied by 100 equals approximately X per cent.
It’s safe to say having X per cent of stock on the market won’t set any alarm bells ringing about Alawa, though the figure wouldn’t have to reach the 80 per cent mark Lawless mentions before investors might want to think heavily about any investments in an area.
Rental vacancy rates
Rental vacancy rates are a vital tool for property investors, Wakelin says, as they provide a reliable insight into the balance of supply and demand for rental property. They therefore provide a guide to how difficult it might be to find a tenant to fill an investment property.
Armstrong says vacancy rates are particularly important if investors are researching regional locations or mining towns for potential investment, as high vacancy rates are a real warning that something’s amiss.
Lomas suggests investors use vacancy rates as a ‘trend’ tool, watching to see whether vacancies are moving higher or lower in a given location over time.
“A downward trend indicates increasing demand,” she notes. “You must independently confirm what’s driving this downward trend to ensure that it’s a sustainable factor.”
Rental vacancy rates measure the proportion of properties available for rent at a given point in time, compared to the total number of established rental properties.
Like data for the amount of stock on the market, seasonal factors have an effect on vacancy rates, as certain times of the year tend to see more rental listings become available. So again, it’s important to compare like-for-like data wherever possible.
API publishes vacancy rates data for postcodes around the country towards the back of each month’s issue in the Databank section of the magazine (this month’s vacancy rates figures start on page 116).
API’s data is sourced from SQM Research; it collates the data by watching online advertisements for rental property over the timeframe of each calendar month. Properties are classified as vacant if they’ve been advertised for two weeks or more and are still currently advertised when the data is being collated.
To calculate the total number of rental dwellings in a postcode, SQM takes the total established dwellings figure from the 2006 Census and then estimates the total dwellings for more recent years. It then multiplies this by the percentage of renters for each postcode, as also provided in the Census.
The vacancy rate is then calculated by dividing the number of vacant rental properties by the total number of rental properties in each postcode.