There are two main drivers of the surge in house prices over the past two decades. The first was the shift to low interest rates. Lower rates enabled Australians to borrow more for a given level of income and so pay each other more for homes. As can be seen, the shift in house prices from below trend to above has gone hand in hand with an increase in the ratio of household debt to income.

The trouble is that the shift to low interest rates occurred in many other countries and most did not have anywhere near the surge in house prices or household debt Australia had, implying a heavy speculative element in driving prices higher as well. I have long thought this surge in household debt and relative house prices represents Australia’s Achilles’ heel. Should anything go wrong with the ability of households to service their debt Australia would be at risk. Fortunately, it’s hard to see the trigger for this in anything but a small way.

The second reason is a lack of supply. While the US saw a property price surge into 2006 matched by a supply surge, supply in Australia has been subdued due to restrictive land supply policies and high stamp duty and infrastructure charges. The National Housing Supply Council estimated a few years ago that since 2001, Australia had a cumulative net shortfall of over 200,000 dwellings. Reflecting this, residential vacancy rates remain relatively low.

Given the supply shortfall, most of the scapegoats that various commentators have come up with to explain high home prices are a sideshow. Foreign and SMSF buying is no doubt playing a role in some areas but looks to be small.

Negative gearing is more contentious, but it’s likely that reducing access, by reducing a tax inducement to investment in property at a time when stamp duty is very high, will have a negative impact on the supply of property. Negative gearing was restricted in 1985, but was reinstated in 1987 after concerns arose that its restriction had worsened the supply of dwellings. Restricting negative gearing for property would also distort the investment market as it would still be available for other investments.

Until we make it easier for builders and developers to bring dwellings to market, the issue of poor affordability will remain.

Rising risks

Our assessment is that the Australian property market is not at the bubble extreme it was at a decade ago: the overvaluation is a bit more modest; annual housing credit growth for owner occupiers and investors is running at around one third the pace seen in 2003; Australians don’t seem to be using their houses as ATMs against which debt can be drawn, suggesting they are less comfortable regarding the outlook and debt; and the home price gains now have been over a shorter period and are concentrated in just Sydney and Melbourne.

However, danger signs are emerging:

- After a cooler period during the first half of the year, the property market seems to be hotting up again. National average home prices rose at an annualised 16.8 per cent pace over the three months to August, according to RP Data, and auction clearance rates are at or above last year’s highs.

- The proportion of housing finance commitments going to investors is now back to around the 50 per cent high seen a decade ago, suggesting that the market is becoming more speculative. And there are signs that homebuyers are starting to extrapolate recent strong price gains into the future, which is very dangerous.

- Finally, The Block, back on top as the most watched show on TV, has highlighted a return to very strong community interest in the property market.

Taken together, these indicators warn that the housing market is getting a bit too hot.

Housing as an investment

Notwithstanding the rising risk of macro prudential controls, in the short term further gains in house prices are likely until the RBA starts to raise interest rates, probably around mid-next year, soon after which another 5 to 10 per cent property price down cycle is likely to start.

Beyond the short term, it’s worth noting that residential property has provided a similar long-term return to Australian shares, with both returning around 11 to 11.5 per cent per annum since the 1920s. They are also complementary to each in terms of risk and liquidity and are lowly correlated. All of which means there is a case for investors to have exposure to both.

At present though, housing looks somewhat less attractive as a medium-term investment. The gross rental yield on housing is around 3.2 per cent and for units it is around 4.4 per cent, giving an average of just 3.8 per cent. After costs, this is just below 2 per cent. Shares and commercial property both offer much higher yields.

Medium-term capital growth is also likely to be limited, with the overvaluation likely to see real house prices stuck in a 10 per cent or so range around a broadly flat trend. This is consistent with the 10-20-year pattern of alternating secular bull and bear phases.

Taken together, this suggests that a realistic expectation for total returns from residential property over the medium term is just around 4 to 5 per cent per annum.

Shane Oliver, chief economist, AMP Capital