We do people use RP Data or the likes of REIWA etc

I often see people using RP Data or REIWA for past history for median prices and so forth.

I know that Australian Bureau of Stats come out a bit slower than the rest but they are a government agency and the Established House Price Index goes back quite a way.

And yes I also use, and have access to, RP Data.

So why are we not using ABS stuff more? Or maybe everybody is and not talking about it.

 

I use both, however ABS stats only drill down to a capital city level so not that useful when looking at a specific area or suburb

 

Also, the ABS stats are sourced from RP Data

 

I have looked at the Established House Price Index but I did not understand it.

Example:

Established House Price Index ; Sydney ; Jun-2016 = 159.7
Established House Price Index ; Perth ; Jun-2016 = 108.8
Established House Price Index ; Adelaide ; Jun-2016 = 114.6

What does this mean?

I couldn't be bothered finding June 2016 medians so I will use August 2016:
Median dwelling price ; Sydney ; Aug-2016 = $780,000
Median dwelling price ; Perth ; Aug-2016 = $480,000
Median dwelling price ; Adelaide ; Aug-2016 = $420,000

Makes a lot more sense to me.

6416.0 - Residential Property Price Indexes: Eight Capital Cities, Jun 2016
House prices jumped again in Sydney and Melbourne during August

To be honest though, I don't pay a lot of attention to Housing indexes or median house prices.

 

An index is only really useful for looking at changes over time. The actual figure at a single point in time is pointless.

For example, if the index at June 2014 is 100, and prices rise 11% over a year, the index at June 2015 will be 111.

Again, if prices then rose 14% from June 2015 over a year, the index will then move to 126.54 at June 2016.

 

Personally I'm pretty dubious about CoreLogic's hedonic index, which they switched to in May (and were shortly after dumped by the RBA). There are several huge problems with their approach.

Firstly, when you hear "house prices have grown 10% this year!" in the media it almost always references CoreLogic. But it includes an average running with their past methodology (stratified median, which APM and the ABS use) and their current one, which is just flat out ridiculous. Doing that risks seriously overstating growth. You can examine their back series and see the huge jump in growth - it's sudden and very marked.

Hedonic indices are rarely used because they compute prices using complex formulae. Essentially, a hedonic index assigns a price based on assuming a value for that house based on sales in the area (even if the house being assigned a value did not sell and hasn't for decades), the number of rooms, size, quality etc. The trouble is, there's a lot of subjectivity - one of a hedonic index's greatest flaws (discussed here http://www.sqmresearch.com.au/RPDailyIndexPiece.pdf).

Secondly, because hedonic indices assign a price regardless of recent sales there's a risk they might understate an early slowdown. If you're assigning a value to a house based on a (partly) subjectively determined metric that includes recent sales in the area, it's hard to capture a stalling market. You can say "oh look this house sold, let's assume all similar houses in the suburb are worth the same" but if a market is sluggish, sales might actually be rare, or vendors might be holding on for a better time to sell, or might have no intention of selling.

Thirdly, a hedonic index is easy to manipulate. A stratified median is hard to mess with because a median is a specific thing. You add up all the sales in a suburb (that's what the 'stratified' bit means) then you find the one smack in the middle and that's your value. It's a little more crude, but over time it gives a good picture. If people are paying more, the median goes up. If people are paying less, the median goes down. A hedonic index on the other hand is a complex calculation that can be tweaked. You can include this or that, change it up, and unless people are going through it with a fine tooth comb it can be hard to call you out if you're making overly favourable assumptions.

Fourthly, CoreLogic's new hedonic index has been moving in (largely) the opposite direction to other major data providers' results. APM just had Sydney's rolling annual growth come in at 2.7% for houses and 0.9% for units/apartments - which reasonably close to the ABS and Residex (repeat sales) results. Not exactly the same, but in the ballpark. CoreLogic's result blows these out of the water. It's got growth sharply (instead of very slightly) accelerating. It's got houses on almost 11% y.o.y (for Oct) and units on over 9% y.o.y - which sounds delightful, but is it realistic?

Anecdotally, my impression of the market is much closer to APM's. In Sydney prices aren't falling but I haven't seen anything much to indicate growth of the strength CoreLogic seems to be suggesting. In many suburbs listings seem to far outweigh recent sales.

So... at least until CoreLogic's hedonic index has been running for a year, it is quite possibly garbage. Even then, I can see why the RBA had serious doubts about it's transparency and reliability. At the very least I'd consider taking their results with a grain of salt, even though their results are the most widely reported by far.