Gap narrows as east almost meets west in housing prices

Most readers should now be familiar with the well-trodden explanation of what factors first drove the unusually sharp house price appreciation in Australia during the 1990s and 2000s. This is a narrative that has been regularly relayed by the central bank and myself.

The short version goes like this: take the past 30 or so years and split them in half around the time that the RBA adopted an “inflation target”. While informally this occurred in 1993, it was officially set in stone in 1996, so let’s take 1995 as a crude line in the sane between the pre- and post-inflation targeting epochs.

Between 1980 and 1995 inflation in Australia averaged 6.6% per annum. Unsurprisingly, variable home loan rates also averaged an astonishingly — for anyone under the age of 35 — high 12.7%. And by 1995, the value of total housing debt held by families divided by their disposable incomes was only about 55%.

Average headline inflation has dropped 58% from its 1980 to 1995 level to just 2.8% per annum in the period since. As night follows day, average variable home loan rates have also plummeted to 7.5%. The 41% decline in home loan costs has, of course, enabled families to boost the amount of mortgage debt they can service for any given level of income. And so Australia’s mortgage debt-to-household income ratio started rising from its 55% level in 1995 to a peak of 156% by 2006. Importantly, it has not increased any further since this pre-GFC apogee — that is, for the past six years, housing credit has expanded at about the same rate as family incomes. Indeed, as at December 2011, it was off a touch at 149.6%, according to the RBA.

With the above insights in mind, our research team has shown that about 96% of the increase in median house prices in Australia between 1985 and today can be explained by changes in household borrowing capacity resulting from the long-term reduction in mortgage rates combined with the growth in family incomes (see first chart below). Said another way, the growth in Aussie housing costs over the past 25-30 years is entirely explicable when one references incomes and interest rates (i.e.we have not even discussed the supply side at this stage).

Yet there appears to be another story to be told. And this one relates to the time that it took for the enormous positive shock to household purchasing power (via lower rates and higher incomes) to become fully capitalised into asset prices.

In the chart above you can see that the blue purchasing power line lies above the red median dwelling price line up until about 2002-2003. This is one way of saying that households could afford to buy more expensive homes — if they assumed the changes to incomes and rates were permanent — but understandably took some time to bid up the value of housing to a level that equated with their purchasing power.

The chart also seems to suggest that, setting aside supply side considerations, asset values did run ahead of income — and interest-rate linked fundamentals in the years after 2002-2003.

In my next exercise I want to lift the lid on the national level data to study how housing costs across states have diverged over time. In order to do so I use statewide median dwelling prices since 1990 from NSW, Victoria, Queensland, Tasmania, Western Australia, South Australia, the ACT and the Northern Territory. The chart below illustrates the change in these raw median prices over time.

While at first it is hard to make head or tail of the chart, if you look closely you can see apparent regime shifts in the statewide capital growth rates: in particular, NSW, the ACT, and Victoria experience very robust dwelling price gains between 2000 and 2003, which, to varying degrees, stops or slows thereafter.House price inflation in Queensland, South Australia, and Tasmania step ups with a lag, but, importantly, continues to kick on after 2003, recording healthy capital growth right through to end 2007.

It is in this period 2003 to 2007 that we also start to find an extraordinary acceleration in the cost of Western Australian homes, which is denoted by the aqua blue line in the chart above. Dwelling prices in the Northern Territory also accelerate alongside Western Australia, albeit at a lower rate.

Here again one sees another difference: while Queensland and Western Australia suffer house price falls during the GFC, which they claw back in 2009 and 2010, home values in the  Northern Territory appear to keep appreciating during the crisis until they finally run out of steam at the end of 2009.

To more precisely measure the time-series divergence in national housing costs I created two benchmarks that are highlighted in the next chart. The blue line is simply the highest state median dwelling price divided by the lowest state median during any given month between 1990 and today.

The red line tackles dispersion another way: it subtracts the difference between the highest and lowest state medians and then divides this number by the national median price. Both metrics tell similar stories.

Irrespective of which measure one uses, the analysis shows us that house prices across Australia underwent an enormous break in the early 2000s fuelled by NSW, Victoria and the ACT, which was only remedied in the years after 2003 by catch-up growth across the other states.

Whereas in the decade 1990 to 1999 the highest state median price was, on average, about 200% the size of the lowest, this differential soared to 321% by the end of 2002. After that climax (refer to the spike in the chart above), the coincidence of much lower capital growth in NSW, Victoria and the ACT combined with faster gains in almost every other region allowed the ratio to normalise back to its pre-boom average of 200%. And, comfortingly, it has stayed around this level since 2006.

One explanation for the phenomenon highlighted might be variations in income growth rates, and hence purchasing power, across the states. Perhaps this was just a case of the NSW, Victorian and ACT economic cycles leading the remaining states.

We can try to address this issue by deflating state house prices by state disposable incomes. Once we have done this, we can repeat the analysis above to determine whether the pattern still holds.

My final two graphs summarise the results over the period 1995 through to end 2011. The first chart shows the individual state house price-to-income ratios indexed to 100 since 1995. The second replicates my earlier analysis by dividing the highest state ratio by the lowest.

The final chart demonstrates that the same dynamics in housing cost dispersion have asserted themselves even after deflating by state incomes: namely, a very large increase in the housing valuation gap from 1995 to 2002-2003, when the highest state price-to-income ratio was about 200% the size of the lowest. By December 2011 this gap had fallen back to just 130%. That is to say, housing costs had converged to their pre-boom levels.

An interesting question for researchers nevertheless remains.  If these changes in housing costs were triggered by the tremendous increase in purchasing power permitted by the decline in mortgage rates in the 1990s, which in turn drove a commensurate increase in household leverage, why did the transmission to house prices  occur at such different times in different states? One for another day, eh?

*Christopher Joye is a leading financial economist and a director of Yellow Brick Road Funds Management and Rismark. The author may have an economic interest in any of the items discussed in this article. These are the author’s personal views and do not represent the opinions of any other individual or institution. This material — first published at Property Observer — is not intended to provide, and should not be relied upon for, investment advice or recommendations.