Australian Property Prices and Data Explained

There are four primary real estate data providers in Australia that publish real estate indices regularly. These are:

  1. Australian Bureau of Statistics (Australia’s official statistical organisation) – publishes quarterly house price movements (see here)
  2. Australian Property Monitors (Fairfax (ASX:FJX), also owns Domain.com.au) – publishes quarterly house and unit price movements and monthly updates (see here)
  3. Residex (bought by Onthehouse Holdings (ASX:OTH) in 2012) – publishes monthly (see here)
  4. RPData-Rismark (a joint-venture between RPData (CoreLogic NYSE:CLGX) and Riskmark (Macquarie ASX:MQG)) – publishes daily and monthly

How are these indices produced?

Real Estate sales prices are sourced directly from real estate agents and from Australia’s various State land title offices who collect information for stamp duty purposes (see list of State authorised resellers in NSW).

This data is used differently by the above companies to determine if prices are rising or falling. Macrobusiness.com.au describes the various methodologies used as follows:

ABS and APM use variants of the stratified median price method, which controls for changes in the composition of properties sold by separating the total sample of properties into a number of geographical sub-samples, and then into different strata based on the long-term average price level of properties in those regions. APM separately measures both capital city house and unit prices, whereas the ABS measures detached house prices only. As the ABS and APM methodologies are closely aligned, they often mirror each other’s house price movements. See here for the ABS’s methodological explanation.

Residex uses a repeat sales methodology, similar to the Case-Shiller Index in the United States, which calculates price changes based upon repeat sales of the same homes. In order to minimise the price-distorting effects that renovations have on home prices, Residex also employs a combination of the hedonic technology and stratified median index technologies used by the other providers.

Finally RP Data-Rismark publishes a hedonic index, which measures price changes for both detached houses and units on a like-for-like basis according to their key attributes, such as location, land size, number of bedrooms and bathrooms, and so on. In March 2012, RP Data-Rismark moved from a monthly index to reporting price changes on a daily non-seasonally adjusted basis, which has led to significant volatility in their results, as well as a question mark over whether it lags the other less frequent indices. You can read more about RP Data-Rismark’s daily index here.

Are the indices accurate?

It is important to understand the limitations of the underlying data. There is a large lag between when the property is sold (exchanged) and when this is reported to the relevant land titles office. This lag means indices are constructed from a relatively small sample set as the majority of the transactions are yet to be reported. APM and the ABS report quarterly to incorporate as much reported data as possible before publication and both revise prior quarter figures (sometimes significantly). RPData-Rismark publishes more frequently – monthly and daily – and has received some negative press of late due to this, particularly from Louis Christopher of SQM Research.

Louis claimed the RPData-Rismark daily index was factually incorrect as he states it was reporting a daily index based on transactions which actually exchanged many months ago (see his claims here and here). His analysis of the sales reported to various land titles offices in May 2013 found that on average they were delayed as follows:

  • NSW 110 days lag
  • QLD 91 days lag
  • VIC 221 days lag
  • ACT 212 days lag
  • WA 113 days lag
  • SA 59 days lag
  • NT 38 days lag

From RPData-Rismark’s response it is clear that their daily index does not directly including past sales into the daily result moreover it factors in where properties were exchanged and this information is used, in addition to exchanges that actually occurred on that day, to form a view of the current market movement.

Which is the best?

Unfortunately there is no clear answer. It all depends on:

  1. The data each provider collects – for example RP Data has a call centre in the Philippines which calls agents around the country on a weekly basis to try and obtain up-to-date sale price data as soon as a transaction occurs. This is used to supplement the data received from the land titles offices.
  2. What they actually use, and
  3. How they use it – the methodologies adopted.

Each methodology has it’s pros and cons. For example from a theoretical standpoint RP Data-Rismark’s hedonic approach is preferred by some of Australia’s leading academic economists such as Professor Robert Hill however it does rely on an accurate data set of individual property characteristics (land size, bedrooms, bathrooms etc). This quality of this information currently available is fairly average, as most agents will tell you.

The somewhat simpler stratified median measures developed by RBA economists (Richards and Prasad) are used by the ABS and APM. This methodology can be adversely impacted by changes to the mix of buyers in the market as well as variations in the quality and type of properties manufactured over time. It is also worth noting that ABS only reports on detached houses in capital cities, APM reporting houses and units in capital cities, and Residex and RP Data-Rismark also reporting regional price changes.

The Reserve Bank of Australia, Australia’s preeminent analytical body primarily references the RP Data-Rismark (see Dwelling Prices here), ABS and APM indices. The indices produced by the Real Estate Institutes and Residex are generally less cited. As macrobusiness.com.au recently showed over time the indices track each other fairly closely.

We suggest viewing each index in light of the above limitations and a blended approach is prudent. From a timing perspective consider RP Data-Rismark’s daily and monthly indices, and supplement this with the ABS and APM quarterly data, which may be more accurate as they incorporate many more sales due to the delay in their reporting.