UNSW Australia UNSW Australia Business School UNSW Australia Business School

House Price Bust: Why Talk of a Bubble Won’t Wash

April 17, 2012

​​​T​he risk of overblown housing prices in Australia has been greatly exaggerated. Historically, economists have worked themselves into a lather about housing bubbles at all the wrong moments. And the recent threat of over-inflated property values in Australia seems to be no exception, according to Nigel Stapledon, an economist  at the Australian School of Business. 

Confusing the issue is the use of several differently measured indices to gauge whether properties are overpriced or undervalued. And the crux of the problem is too little information.

Australia’s residential property market defied global trends by avoiding the collapses suffered by other countries during the global downturn. Insulated by a strong local economy, low unemployment, a growing population and government schemes that propped up the first homebuyers’ market, Australian house prices slipped only 3% in 2008 as the US and UK recorded double-digit slumps.
However, when it comes to looking at where Australian property prices are headed, opinion is diverse, ranging from gloomy pessimism to almost insane optimism. Meanwhile, Australian house prices have been steadily falling.
According to property information firm RP Data-Rismark, seasonally adjusted home prices dropped 4% in 2011 and are forecast to slide further despite two interest rate cuts late last year. Overall, distressed real estate listings for the country rose 30% in 2011.
Since the beginning of this year, auction clearance rates in Sydney have averaged about 55% – notably higher than the rates recorded at the end of 2011, when agents struggled to sell 50% of the properties being put to auction under the hammer – but nowhere near the high 60s and low 70s that were achieved in Sydney during times of past house price growth.
Will this recent downturn in the housing market continue? Or, is Australia effectively in the grip of a housing bubble that may burst at any moment? Bearish property analysts swear house prices need to drop 40% to return to their long-term levels and those comparable countries. They suggest the property market would have collapsed in 2008 if it were not for first-time buyers being enticed to dive into the market by inflated government grants.

Yet, others dismiss the idea of a bubble as nonsense, claiming prices are more likely to flatline than crash. They see any major slump as unlikely given the strong economy and Australia’s “chronic” housing shortage. According to the National Housing Supply Council (NHSC), the estimated dwelling gap for June 2009 was 178,400 and was expected to increase by 72.6% to 308,000 by 2014.
This anti-bubble theorists view softer house prices as an indication that people have simply quit speculating and are becoming more debt-averse.
Philip Soos, a researcher at Deakin University, is in the pro-crash camp. He notes that eight of the nine substantial property price increases over the last 131 years have resulted in a decline. That being the case, he can’t see why the largest increase on record would not precede another drop now. Soos does not buy the argument that Australia is suffering from a “chronic housing shortage”, saying property prices have continually experienced boom-bust cycles regardless of population growth levels. “Housing prices started to rise in 1996 and skyrocketed from 2001 onwards. Yet, 2007 was the first time since 1951 when population growth exceeded dwelling growth, so housing prices should have started to rise from 2007 onwards, not 1996,” Soos points out. Once a bubble bursts, he notes, a huge oversupply of housing is likely to emerge.
Soos argues the rapid rise in residential property prices cannot be explained by economic fundamentals, including the downturn in inflation and interest rates, which began in the late 1990s. “If lower inflation and interest rates were enough to cause a run-up in prices, the 1960s should have prompted an even greater level of household debt. Yet housing prices and rates were the lowest on record. Currently, while nominal rates have fallen by half, the amount of mortgage debt has more than quadrupled, indicating an excessive amount of debt.”
Lax lending standards and that fact that land is relatively untaxed have fuelled the run-up in prices, says Soos. “The obvious indicator of a bubble driven by speculation is a large gap between rent and home sales prices – which is the case almost everywhere in the West where property overvaluation has occurred, including Australia.”

Measure for Measure

Nigel Stapledon, an economist at the Australian School of Business, agrees with Soos on the overuse of the population growth argument, but he does not believe house prices have somehow disentangled from the fundamentals. He has another explanation for why there is such dissent over property.
Since the early 2000s, many economists have been using house price-to-rent ratios to gauge whether or not home prices are inflated or undervalued. The ratio compares house prices with rental indices. For the ratio to work, these indices should capture identical data. But they don’t.
Many house price indices use raw price data whereas the Australian Bureau of Statistics rental index adjusts for quality improvements made over time to a property. A house sales series with no adjustments for alterations and additions, which have boosted the size of dwellings over time, will tend to rise faster than a rent series.
The two indices are measured very differently resulting in a biased measure, says Stapledon. “If you don't allow for the measurement issue, you get a dramatic upward trajectory in the price-to-rent ratio and people see a bubble,” he says, “Standardise the indices and that upward slope changes dramatically from an upward trend to a flatter trend, in Australia’s case.”
Stapledon has attempted to fix this “apples and oranges” problem by creating an index that uses a rental income series to match the house price series. Both series incorporate quality changes such as additions and alterations, thus removing the bias. While there’s a divergence between rents and prices, Stapledon reports the scenario is not as scary in 2012 as it appeared in 2003 when soaring prices had pushed the price-to-rent ratio into bubble territory and the market was vulnerable to a sharp correction.
A bubble implies prices are out of line with fundamentals. In Stapledon’s view, that initial rise in the price-to-rent ratio from 1996 was a lagged response to the earlier decline in interest rates. Stapledon’s analysis differs from Soos’. Stapledon thinks that after the pain of high interest rates in the 1980s, which soared to more than 17%, it took some time for borrowers to believe that rates would stay down. That’s why there was a reluctance to take on debt.
Economists have never been very precise at predicting the timing and magnitude of corrections, notes Stapledon. Through the late 1990s, economists talked about a housing bubble – yet the market rose by 50%. Outstandingly, most US economists did not see their enormous housing bubble until it burst in front of them. But Stapledon insists many analysts in the “bubble” camp have not looked closely enough at the Australian situation. “The reasoning is if the US has a housing bust; then so too should Australia,” he says. “This reasoning fails to appreciate the very powerful influence of the resources boom. Australia is not alone in this. There has been no bust in Canada, which has also experienced a resources boom.”

Not Enough Information?

Soos is concerned there is too little data in the public domain, a view shared by Michael Sherris, a professor of Actuarial Studies at the Australian School of Business and chief investigator in the ARC Centre of Excellence in Population Ageing Research (CEPAR). Furthermore, the publicly available data can be hard to follow. “It is very important that we have better data and that indexes are reliable,” says Sherris.
One problem is that indices for the broader market do not provide information for an individual house relative to others, leaving homeowners without insights on the market value of one of their biggest assets. For example, the sale price of a two-bedroom house with a garage in one area cannot be extrapolated to another house in a different area in a consistent and reliable manner based only on market indices.
“We need better more refined publicly available indices to understand what’s going on in the residential property market allowing for how house prices vary according to their characteristics,” says Sherris.
While privately collected and analysed data might eventually be dispersed to private customers such as banks, the public remains largely unable to access the analytics around individual house price characteristics and risks, which means the available data – regardless of how well it is analysed – is less than helpful.
RP Data-Rismark’s Ben Skilbeck says his firm overcomes the problem by using "hedonic'' methodology, which takes into account the characteristics of homes bought and sold. Typically, this includes the number and size of rooms, the number of bathrooms and how far the residence is from a major road and schools, along with special features such as swimming pools and tennis courts. The hedonic model uses such variables to understand their relative value contribution to a property. Once this is understood, changes in the composition of houses selling in one period relative to another are controlled, enabling the underlying market movement to be determined.
The firm tracks the rental market in a similar way. This eliminates the “apples and oranges” problem that exists for many reported rental yield numbers, claims Skilbeck. Median advertised rents are commonly divided by the median house sale prices to quote rental yields, he says, which leads to inappropriate conclusions because the rental houses are often very different to the properties sold.

Spot the Difference

The problem with using a median price to calculate a housing price index is that the composition of houses sold in one quarter may differ to those sold in another, indicates Skilbeck. “Suppose that over a particular period, there were more sales of larger houses in more affluent suburbs. Simply looking at the sale prices would give an inflated view of market movements. Conversely, if in the next period, there were more sales of smaller houses in less affluent suburbs, looking at only the sale prices would lead us to believe market returns were lower than reality.”
Skilbeck points to the increased activity in the property market in early 2009 when the federal government doubled the first homeowners’ grant and state governments provided various first home buyers with stamp duty concessions and grants. At the same time, the central bank, the Reserve Bank of Australia, slashed official interest rates by nearly 4% between September 2008 and April 2009 from 7% to 3.75%. The disproportionate number of lower-end properties selling during that period artificially dragged the median house price downwards when in reality the market was responding strongly to the economic stimulus. When the grants were phased out and first homeowner activity decreased dramatically, the median indices showed the market was rising because fewer lower-end first homeowner properties and more higher-end properties were transacting, but the market was actually pulling back. “Our indices showed the true state of the market, whereas the figures from the Australian Bureau of Statistics simply showed a change in the composition of the type of houses being bought and sold,” says Skilbeck.
Trying to gauge the sales price of houses based on bedrooms, swimming pools, views and other commonly valued characteristics is not easy.
Louis Christopher, head of Sydney-based property research and advisory company, SQM Research, warns of inherent weaknesses in a model that tries to measure house price movements by taking into account quality differentials between various homes. “Overall, it is a good theory, but in practice it still has its flaws,” he says. For a hedonic index to work, it needs a vast amount of data on individual properties.” The index does not look at all housing features such as the slope or the shape of the land, renovations or improvements, which are major contributors to value, as Stapledon’s research shows.
Valuing a portfolio of assets, which are all different and most of which do not trade in any period, is a complex problem. “No method can exactly estimate the true value of every property. Even professional valuers can’t do it,” concludes Skilbeck.​
comments powered by Disqus

Subscribe now

BusinessThink is a free online publication. By subscribing, the latest edition will be delivered to your inbox once a month.

​ ​
Print PDF