The final word on Australian house prices: Christopher Joye

The final word on Australian house prices: Christopher Joye
Christopher JoyeDecember 8, 2020

There is an enormous amount of time wasted debating whether Australian house prices are over- or under-valued. The vast bulk of everything that is written on this subject is, unfortunately, both factually flawed and misleading. 

Today I want to objectively eviscerate the long-standing myth that the Australian housing market is in the throes of an unsustainable bubble using two very simple charts, which you are unlikely to have seen before. 

By referencing this analysis alone one can conclude that current Australian house prices are comfortably explained by fundamental economic factors. That is, there is no evidence that domestic housing costs have overshot basic measures of fair value. We will also see that the average Australian family’s disposable income after meeting all housing costs is actually higher than it has been at any time since our analysis begins in 1985. 

The first question we address is most powerful in its simplicity. Specifically, we want to work out how much of the growth in Australian house prices since 1985 can be explained by changes in disposable household incomes and mortgage rates. 

Regular readers will recall that both the RBA and I have regularly argued that the increase in household debt-to-income ratios, which started during the early 1990s, and the coincident rise in Australia’s dwelling price-to-disposable income ratio, can be partly attributed to the circa 40% reduction in average mortgage rates between the 1980 and 1995, on the one hand, and 1995 and today, on the other. These two periods book-end the introduction of the RBA’s “inflation-targeting” regime, which helped lock-in consumer price growth of 2% to 3% per annum, and thus much lower nominal interest rates. More formally, mortgage rates averaged 12.6% between 1980 and 1995. Since that time, mortgage rates have averaged a little less than 7.5%. 

In our first chart below Rismark’s research group has put this hypothesis to the test. In particular, we take the median Australian dwelling price in June 1985 and “index” this by only two variables: (1) changes in incomes; and (2) changes in borrowing capacity enabled by variations in mortgage rates over time (assuming that the household’s loan-to-value ratio and mortgage repayments-to-disposable income ratio stay constant in 1985 terms). 

The final income- and interest rate-adjusted 1985 median dwelling price is denoted by the blue line in the chart below. The red line represents actual median prices. 

To get the best possible measure of the change in household earnings, we use the ABS’s disposable household income series, which is reported as part of the National Accounts. We have not used the gross increase in disposable household income, since this is artificially inflated by population growth. To control for the latter, we divide disposable household incomes by the number of households each quarter. This gives us disposable income on a per household basis, which expands at a lower rate than the headline disposable income series. 

If we simply index the 1985 median dwelling price by this measure of incomes, we can account for 60.1% of the total increase in Australian housing costs over the last 26 years.

Click to enlarge

As discussed, the second variable we use to adjust the 1985 median dwelling price is the change in household borrowing capacity over time. To properly quantify this, we keep the amount of money households spend on their mortgage repayments constant. That is, we fix the 1985 mortgage repayments-to-disposable income ratio and apply this fixed ratio to the entire 1985 to 2011 time-series. 

We do not, therefore, allow households to spend any more of their earnings on meeting housing costs over time. We also assume a constant mortgage loan-to-property value ratio. Accordingly, the only variable that influences the household’s borrowing capacity is fluctuations in mortgage rates over time, which we source from the RBA. 

Adding changes in borrowing capacity to the blue line in our chart above, we can explain a further 31.9% of the increase in Australian housing costs since 1985. 

Collectively, incomes and interest rates account for 92% of the increase in median Australian dwelling prices over the period 1985 to 2011. 

Note that since the average size of Australian homes has tended to rise, it is likely that there are some “compositional biases” affecting our median measure of housing costs (i.e., the red line in the chart above). That is, it is conceivable that incomes and mortgage rates actually account for an even higher share of the time-series change in dwelling prices than our analysis suggests. 

Our second chart takes a novel look at the question of housing affordability. This analysis actually borrows from an approach first developed by Dr Tony Richards at the RBA. Specifically, it asks the question: how much real disposable do households have left over after buying the representative home and servicing all their mortgage costs? 

In order to compare the welfare of different households over time, we take the disposable income measure used in the first chart above every quarter and assume the household buys the median-priced dwelling as recorded in that same quarter. Given a fixed loan-to-value ratio, we then work out the principal and interest repayments owing on the household’s mortgage, which it meets out of its income. After adjusting for inflation we have a quarterly estimate of real disposable income per household netting out all housing costs (i.e., after buying a home and servicing the mortgage). This is represented by the blue line in the chart below.

Click to enlarge

It turns out that in 2011 Australian households actually have more income left over after purchasing the dwelling of their dreams and meeting their principal and interest repayments than any of their predecessors since our analysis begins in June 1985. This was exactly the same result that Richards found when he first published his work in early 2008. On the basis of this benchmark, housing affordability in Australia has never been better.

In closing today’s column, I want to provide an update to our closely followed dwelling price-to-income ratio. We have recalculated this data to reflect the ABS’s September National Accounts. The key conclusion is, as we have predicted for some time, that Australia’s dwelling price-to-income ratio continues to slowly decline to around four times, which also accords with what the RBA finds.

While one can quibble about exactly what the ABS includes in its measure of disposable incomes, we are more concerned about changes in the ratio over time (rather than the specific levels). And when focussing on how the ratio has moved, we can clearly see that there is no evidence that Australian housing costs have grown more rapidly than disposable household incomes over the last seven to eight years. In fact, when we use a much more precise measure of housing costs — namely, our hedonic indices – we find that per capita incomes in Australia have actually outpaced national dwelling prices by about 15% since the end of 2003.

Click to enlarge

In a recent report, the ratings agency Moody’s downgraded its outlook for Australia’s mortgage insurance sector almost exclusively on the basis of concerns about Australia’s housing market. On the one hand, Moody’s argued that, “elevated house prices and mortgage debt levels are arguably at unsustainable levels and remain a key vulnerability for the LMI industry.” Yet in the same report they concede that “our view is that the data on the sustainability of Australian housing prices remain ambiguous.” While these two statements are difficult to reconcile, we hope that the evidence presented above will help clarify Moody’s views.

 Christopher Joye is a leading financial economist and a director of Rismark International and Yellow Brick Road Funds Management. The above article is not investment advice.

Editor's Picks