I was surprised by the statement made by Westpac's Chief Economist today
It is impossible to tell what is really going on with house prices
His comments refer to the growing volatility in house price measures, especially from the latest REINZ stats of both median price and stratified median price index. The latter a measure designed to provide a more accurate and stable measure of prices as it balances out the impact of high price suburbs vs. low price suburbs to ensure neither unduly skews the data.
This chart above shows the trend in both median and stratified price over the past 7 years - most noticeable is the rise of the recent 3 years. You can also see the volatility in recent months from what has historically been a smoother trend in stratified price measure as seen in the red line.
In the Auckland market the volatility is more pronounced and also what is very striking is the direct correlation of prices (especially the median price) with the timeline of the implementation of the Reserve Bank LVR restriction.
The impact after the 1st October implementation has seen median price shoot up directly as a result of the significant fall in sales of lower priced property thereby pushing the median price up. The fall in January prices is a seasonal issue which can be seen regularly through the past years, although the scale of the fall this year is surprising.
As Westpac's Chief Economist Dominick Stephens says in his Home Truths report for April the fundamental issue is that the raw data of house prices is not a case of matching apples with apples, more its an apples with oranges comparison. On the REINZ Stratified House Price Index (HPI) he makes these comments reflective of the charts above:
As I have written in the past I believe that REINZ should take on the challenge of collecting and aggregating more granular statistics on property sales - they are in the best position to do so as they are the incorporated society with almost all real estate agents as members who collect data on every sale in NZ at the time the unconditional agreement is reached.
Even without changing any collection process they could today provide far more accurate and valuable information on property prices.
One of the fundamental problems with there data reporting today is that they currently aggregate all property types together - be it a 1 bedroom unit, a studio apartment, a 5 bedroom home or a 20 hectare lifestyle property - to the Real Estate Institute they are all just a house - that is dumb!
Take a look at the divergent make up of property in Auckland across the major districts as defined by REINZ as seen from the composition of listings on Realestate.co.nz today.
Certainly across Auckland two thirds of properties for sale are houses, but within Auckland city over 1 in 3 of all properties are apartments. One in ten of all properties on the market across the Auckland region are units or townhouses - these certainly have a very different price profile than regular houses or lifestyle properties. In the outer areas of Auckland Lifestyle properties make up more than a quarter of properties. Think on the fact that within Manukau a $4m lifestyle property in Whitford is seen by REINZ as the 'same' as a $219,000 single bedroom unit in Papatoetoe.
All sales records remitted to the data system at REINZ currently have the following fields (some of which are completed by the selling agent - certainly all should be mandatory)
- Unit Number
- House Number
- Road Name
- List Price
- Sale Price
- List date
- Agreement date
- Unconditional date
- Type - Residential House / Apartment / Home & Income / Unit / Lifestyle
- Land Area
- Floor Area
- Valuation Year
This is the data set for existing data - upon this set of data, better more accurate sales and sale price analysis could be undertaken to allow economists, property buyers, property sellers, investors, real estate agents and other could make better informed decisions.
Maybe then we could avoid the statement that an Economist feels that its virtually impossible to know what is going on with property prices.