Yesterday it was widely reported that housing affordability had increased to its highest level in 12 years. Quite frankly, that result is a mirage, supported by dubious data, and completely at odds with the experience felt by many Australians.
On a quarterly basis, the Housing Industry Association releases a housing affordability measure which assesses whether it has become easier or harder for Australians to service their mortgage debt.
Yesterday we reported that this index had improved by 2.1 per cent in the March quarter. It is a curious result given house price growth has been fairly strong -- particularly in Sydney and Melbourne -- while wage growth is at its lowest level in 16 years. In fact, once inflation is taken into account, real wages have actually begun to decline.
Compounding matters, both standard and discounted variable interest rates have remained unchanged since July last year. Recently the three-year fixed rate has declined modestly on lower expectations for the cash rate, although this interest rate is not considered in the HIA housing affordability measure.
What’s going on?
Well, the improvement in affordability is mostly due to a fairly dubious measure of house prices.
The median dwelling price measure used by the HIA is based on the Commonwealth Bank’s loan approval figures for that quarter. It is a simple median measure -- in contrast to other measures of dwelling prices -- which means that it does not take into account changes in the mix of size, location or quality of the dwellings financed.
This might not sound like a big deal, but dwelling price measures that fail to adjust for compositional change are typically excessively volatile and do a poor job of measuring dwelling price growth.
For example, following the onset of the global financial crisis, the federal and state governments implemented a range of very generous grants for first home buyers. This triggered a sharp rise in first home owner activity, which can be seen in the graph below.
For a simple median measure of dwelling prices, a sharp rise in first home owner activity could actually cause dwelling prices to fall because first home buyers typically buy houses or units that are cheaper than average. This occurred despite the fact that government grants were bidding up prices significantly over that period.
A good dwelling price measure will try to compare like-for-like. The RP Data measure, for example, uses a hedonic modelling technique based on a range of variables (such as location, number of bedrooms etc) to control for changes in the type of dwellings sold each month.
The measures from the ABS and Australian Property Monitors (APM) utilise a stratified median approach. This involves separating each capital city into postcodes and then allocating those postcodes to ten buckets determined by the median price in each postcode. They take the median price for each bucket based on transactions during the period, and apply a geometrical mean to create a median price measure for the entire city.
Both measures might sound complicated but they do a very good job of reducing monthly or quarterly volatility, ensuring that price growth mainly reflects actual changes in prices rather than changes in the type of houses or units sold.
The Commonwealth Bank measure does no such thing, and it’s not surprising that it arrives at unusual results. According to this measure, dwelling prices have declined by 2.1 per cent in the March quarter, to be 1.8 per cent lower over the year. I’m sure that would come as quite the surprise for residents of Sydney or Melbourne.
By comparison, the ABS measure increased by 1.7 per cent in the March quarter to be 10.9 per cent higher over the year.
Housing affordability or mortgage affordability has surely improved over the past couple of years as mortgage lending rates have declined. But the sharp fall in first home buyer activity suggests that housing remains unaffordable for many Australians.
The HIA result reported yesterday is completely illogical and reflect a dubious measure of dwelling prices. It turns out it is very easy to make housing affordable if you use imaginary data, but that won’t help the average Australian afford a home.