Property investors need to realise that widely quoted industry analysts employ very different techniques for measuring property values.
THERE has been some sensational, rather than balanced, reporting of property market trends recently.
This occurred particularly during the global financial crisis, when many reports claimed that Australian residential property prices were likely to crash. Now the press is providing sensationalist comment in the other direction, focusing on the recent report by the Australian Bureau of Statistics that property prices have increased 20 per cent during the past year. We believe that the actual increase is far more modest.
The table (top right) shows the growth in house prices over the past year estimated by four separate sources.
A rise over the year of 10 per cent (Residex) is very different indeed to a rise of 20 per cent (ABS). And the quarterly difference is even more extreme.
It is therefore important not to rely on the headlines, but to undertake a degree of research and look behind the headlines.
To make sense of conflicting figures we need to understand the different ways house prices are calculated.
No measure can be perfect, because the dwellings sold in each period are different. This means that various methods of estimation must be used, each with their different flaws.
Many measures use some concept of the ‘median’ price, calculated by ranking all sales from highest to lowest, and taking the price of the dwelling in the middle of the sample.
A problem with the median price is that even if the price of each individual dwelling shows no change at all, the median price can still change.
The median dwelling – the one which falls in the middle of the range – will be:
• a cheaper dwelling if activity is biased toward cheaper dwellings, as occurred a year ago when there was a lot of activity in cheaper dwellings from first home owners; and
• more expensive dwelling when there is more activity in expensive dwellings, as there is now.
In this example, the change in the price of the median dwelling has nothing at all to do with a change in dwelling prices – it is purely a result of high activity in one part of the range altering the type of dwelling that falls in the middle of that range.
Applying this to the current situation, we see that the latest median figures compare a market where the proportion of cheap homes sold was low (now), with a period when it was high (a year ago) – it is no surprise that the two medians are well apart, because the groups of sales being compared are quite different.
This change in activity of first homebuyers is evidenced in the RBA chart (below right).
The other interesting feature of the chart is that investors are moving back into the market.
• Do not to take press reports, or even expert analysis, at face value. Take the time to understand how the various price estimates are calculated, so that you can interpret them more sensibly.
• Recognise that property investment is a medium/long-term proposition – over the longer term, the various measures fall into general alignment. Make any investment decision based on long term fundamental forces.
The main measures used to calculate dwelling prices are outlined in general terms below.
The Real Estate Institute
The REI uses a simple median – the sales are ranked from highest to lowest and the one in the middle is picked. (In a statement issued last week, which reported 3 per cent growth in Perth’s median house price, REIWA acknowledged that, “much of the rise in median price reflected compositional changes in the market and not necessarily increased values”.)
Australian Bureau of Statistics
The ABS has in recent years upgraded its approach to a ‘stratified median’. It divides (or stratifies) each capital city into clusters, based on similarities in the type of dwelling – there are five regions in Hobart and 22 in Sydney. It takes the median of each cluster. They it weights each median by the estimated importance of its cluster based on 2006 census data.
This approach will help to eliminate flaws caused by a change in activity between the various clusters, but not within the cluster itself – a shift of activity within each cluster from low- to high-quality dwellings will bias the figures.
The ABS sample data do not include units or semi-detached dwellings (about 25 per cent of dwellings in capitals) or dwellings in regional areas (about 40 per cent of dwellings).
Australian Property Monitors
We understand that APM also uses a ‘stratified median’ approach, similar to the ABS. However, we understand that fewer strata are used and that the strata are not weighted. APM calculates capital city prices only, separately for both for houses and units.
Residex use a ‘repeat sales’ methodology. This looks at the change in price from properties that have been sold before.
This approach does not suffer from the shortcomings of the median or stratified median approach. But it has different shortcomings – it excludes dwellings where data on earlier sales is not available including new dwellings, and it excludes the value of renovations.
Residex has house and unit data for each postcode region of Australia.
RP Data-Rismark is the most recent producer of dwelling price indexes. The most sophisticated of their indexes is an ‘hedonic’ index, which attempts to make adjustments for the unique attributes of each property, including location, landsize, type (eg house or unit), number of bedrooms and bathrooms, etc.
Not all statisticians support the use of a hedonic index. For example, the ABS states that: “Sufficient data is not available for timely production of the index using hedonic methods. It would be necessary to collect considerable amounts of supplementary data. Due to the cost of collecting and processing this data, the ABS considers that the hedonic approach is not viable at this time.”
It is indeed expensive. RP Data spends more than $9 million annually collecting new property information and has amassed a database comprising more than 140 million property data records covering 99 per cent of dwellings. Its hedonic index has received supporting comment from representatives of CommSec, Macquarie Bank, the Securities Industry Research Centre of Asia Pacific, and Moodys.
RP Data-Rismark includes all regions, house and unit data separately, and seasonally adjust the figures to take account of the fact that certain quarters typically show stronger growth than others.
In my view, in a period in which there have been unusual levels of activity by first homeowners, the median approaches are likely to have a bias, and the approaches of RP Data Rismark (hedonic index) or Residex (repeat sales) are likely to be more accurate.
This suggests that growth in dwelling prices is not nearly as high as the recent headlines have suggested.
That is not bad news for investors. It means there is scope for continued steady growth in dwelling prices.
History tells us that the current environment has characteristics that lead to sustained growth in dwelling prices: recovery from a share market crash; recovery from an economic downturn; strong economic growth; strong population growth; and, importantly, a huge undersupply of residential dwellings.
Rising interest rates obviously weigh against rising dwelling prices. However, historically there is no statistical correlation between interest rates and dwelling prices (often they rise together) – this is because interest rates are usually raised to try and lean against the forces that are pushing property prices higher and are only partly successful.
Clearly, the past year has been a good time to buy well-researched investment property. However, in my view, the overall cycle still has a considerable way to run and the next few months will continue to provide excellent buying opportunities for well-researched property that is supported by strong economic fundamentals.
Michael Shreeve is a director of Compass Capital Property Group, which provides property investment opportunities to accountants and financial planners.