The Great Australian Housing Bubble

 |  Includes: ADE, EWA, FXA, KROO
by: Leith van Onselen

There is currently a widespread debate in Australia and abroad over whether Australia is experiencing an unsustainable housing bubble, or asset inflation based upon sound fundamentals.

Recent analysis by the Economist, the IMF, Demographia, Morgan Stanley, and Jeremy Grantham of GMO argue that Australian house prices are severely overvalued and due for correction.

On the other hand, Australia’s banks, the property industry, and industry 'experts' claim that Australia’s housing market is different to other countries and underpinned by sound fundamentals, including a strong economy, high population growth and housing shortages, as well as a robust banking system.

So which camp is right? Is Australia’s housing market a debt-fuelled time bomb underpinned by ‘Ponzi finance’ that requires the ‘greater fool’ and ever-increasing levels of debt to perpetuate it? Or are current valuations justified? Let’s examine the data.

One long boom:

The below chart plots Australia’s nominal house price growth against comparable countries. As you can see, Australia’s house prices growth has been nothing short of phenomenal, leaving the other countries in its wake.

Click to enlarge:

Click to enlarge

Next I have plotted average Australian established house prices against average Household Disposable Incomes (HDI) and Average Full-Time Ordinary Earnings (AFTOE).

Click to enlarge:
Click to enlarge

As you can see, the ratio of house prices to average earnings started at around 2.5 times HDI and 3.7 times AFTOE in 1986. This ratio increased slowly from the mid-1980s to 2000, rose rapidly from 2000 to around 2004 and then settled at around 6 times HDI and 7.7 times AFTOE in 2008/09.

The increased cost of housing is also reflected in the amount of HDI required to service the average mortgage (see below chart).

Click to enlarge:
Click to enlarge

As shown above, despite as significant reduction of mortgage interest rates over the past 20 years (shown below), the ratio of average mortgage interest payments to average HDI has increased from around 6 per cent in 2000 to 10 per cent currently (after peaking at around 11.5 per cent in 2008). Whilst this ratio may look benign on the surface, only 35 per cent of households in Australia have a mortgage. So the actual repayment burden on indebted households is much larger than implied by the average.

Click to enlarge:
Click to enlarge

While commentators can argue about the choice of house price data and income measures, the trends are clear – Australian housing has become far more expensive over time and households are now required to dedicate a much larger proportion of their lifetime’s earnings to purchase a home.

Debt levels explode:

Since the growth of house prices has significantly outpaced the growth of incomes, it follows that rising debt levels have been the key contributor to rising house prices in Australia, since the only way to purchase something that you cannot afford through income (or savings) is to borrow the difference. The below chart uses RBA data to plot the level of mortgage debt against HDI and GDP.

Click to enlarge:
Click to enlarge

As you can see, Australian mortgage debt has increased significantly from 32 per cent of HDI and 12 per cent of GDP in 1990 to 142 per cent and 92 per cent respectively as at June 2010.

Further, Australia’s household debt levels - which includes mortgage as well as other personal debt - compares poorly against other advanced nations (see below chart).

Click to enlarge:
Click to enlarge

Based on the above data, we can confidently conclude that Australia’s house price growth has been debt-fuelled, thereby fulfilling one of the key ingredients of an asset bubble.

Ponzi buys a house:

What about the claim that Australia’s housing market is being underpinned by ‘Ponzi finance’, whereby the rental income (or imputed rent) from a property does not cover the debt expense incurred to purchase the asset. Then, therefore, requiring perpetual capital growth, the ‘greater fool’ and ever-increasing levels of debt to perpetuate it?

To evaluate this claim, I have first plotted the growth in real (inflation-adjusted) house prices against the growth in real rents. This allows us to assess whether rents have increased at roughly the same pace as house prices such that rental incomes broadly cover the cost of debt repayments.

Click to enlarge:
Click to enlarge

According to the Australian Bureau of Statistics (ABS), real rents have increased by around 15% since 1987 while real house prices have risen by around 165% over the same period. It is no surprise, then, that yields on rental houses have plummeted from around 8% in 1987 to 3.5% currently (see below).

Click to enlarge:
Click to enlarge

Remember that the rental yields shown above are before deductions for property expenses such as rates, land tax, maintenance and agents fees. If you take these costs into account, then current net rental yields would likely track below 3% - far below the current discount variable mortgage rate of around 7%. Put another way, the average new housing investor would incur a pre-tax income loss of over 4% on every dollar invested in housing.

Yet despite the worsening investment fundamentals, investors have piled into residential property at an astonishing rate. The below charts, which come from Morgan Stanley, shows that Australia has become a nation of landlords. In 1988/89, 608,000 taxpayers reported rental income to the Australian Taxation Office (ATO). But by 2007/08, 1,765,000 taxpayers did – 13% of the total.

Part of this surge of housing investment can be explained by the Baby Boomer generation reaching peak earnings age (45 to 55 years) from 1990. As the Boomers and others realised that they had not saved enough for their retirement, they began buying investment properties en masse as a way of catching up in a hurry, helped along by a proliferation of tacky property investment seminars marketing slogans like: “THIS WEEKEND CAN MAKE YOU A MILLIONAIRE” and continuous segments on tabloid television showing every man and his dog making fortunes on the back of property.

But with house prices booming and rents remaining flat, the proportion of taxpayers where rental income does not cover costs increased from around 50% in 1993/94 to 70% in 2007/08:

Meanwhile, the aggregate level of net rental losses has ballooned from +$219m in 1999/00 to -$8,600m:

Finally, as shown by the below ATO table, the overwhelming majority of loss-making investors are middle income earners, with 77% of these investors earning less than $75,000 per year in 2007/08. By comparison, only 5% of loss-making investors earned over $150,000.

Click to enlarge:
Click to enlarge

The data, therefore, strongly suggests that the Australian housing market is being underpinned by Ponzi finance, whereby investors and owner occupiers are leveraging up to buy property in the hope of achieving continued rapid capital growth (in the case of investors) or ‘getting in’ before prices increase further (in the case of owner occupiers). But with the significant negative income returns (or imputed rents) from holding residential property, the only way that house prices can continue to increase faster than incomes is if buyers believe that prices will continue rising and that large capital gains can be made by selling the same asset to other buyers (the ‘greater fool’ theory). Such a scenario also requires ever-increasing debt levels, which is clearly unsustainable.

This hypothesis is broadly supported by a recent investigation by the Economist, which found Australia’s housing market to be the most overvalued in a sample of 20 countries using an average price-to-rents methodology. Similarly, the IMF recently found the Australian housing market to be amongst the most overvalued in the OECD based on price-to-rents and price-to-incomes.

Risky finance:

While there are many factors that have increased the demand for housing – such as Australia's generous taxation system (explained here), subsidies paid to first-time buyers, and Baby Boomer demographics – the extra demand for housing could only have fed into higher prices if credit was readily available, enabling buyers to borrow large sums and pay high prices. Put simply, the supply of credit is the crucial ingredient to Australia's booming housing market.

As explained in the book, The Great Crash of 2008, it was the rise of the non-bank lender in the mid- 1990s - raising funds via securitisation activities on the wholesale debt markets - that initially caused an intensification of competition amongst mortgage lenders. These non-bank lenders have no formal regulator and no rules outside of regular trade practices and corporations law. It was they who led the decline in Australian credit standards from the mid 1990s by introducing ‘innovative’ loan products like low-doc loans in 1997, then ‘no-doc’ loans in 1999. More recently they were beginning to issue ‘non-conforming’ (sub-prime) loans just before the global recession intervened.

Faced with this new competitive threat, Australia's banks responded in kind by reducing their deposit requirements and tapping new sources of funding offshore. Gone were the days of requiring a minimum 20% down payment and banks funding their loan portfolios from domestically sourced funds (mostly deposits). Instead, 5% down payments became commonplace, funded increasingly by the banks issuing bonds to foreigners.

As shown below, the percentage of bank liabilities funded from foreigners has increased from just over 5% in 1989 to around 22% currently, totalling over $500 billion! Over the same period, the banks increased the proportion of loans channelled into housing, with housing loans increasing from around 35% of total lending in 1990 to 57% currently.

Click to enlarge:

Click to enlarge

Anyone seeking an answer as to why Australia owes so much money to foreigners only has to look to the banks' model of borrowing heavily offshore to pump-up housing: (Click to enlarge)

Click to enlarge

Banks are awash with funds - sourced from both domestic and foreign sources. With a higher proportion of bank assets (loans) being directed into housing, is it really a surprise that house prices and household debt levels have exploded over the past two decades?

The key risk is that Australia’s ability to sustain current house prices, let alone further price increases, rests with the willingness of other countries to continue lending Australia's banks money. But in times of crisis, such as when Lehman Brothers collapsed, foreigners tend to zip-up their wallets. The crisis left our banks, house prices and broader economy exposed to a sudden liquidity shock as the banks were unable to roll-over their foreign borrowings (let alone increase them).

Few people realize that the Australian Government’s October 2008 guarantee of bank funding and deposits was issued after the larger banks made it clear to the Government that they were facing extreme difficulty in rolling over their wholesale funding, meaning that they would have to immediately withdraw credit from the Australian economy and might eventually face insolvency. So while it might be true that Australia’s banks satisfactorily managed credit risk, avoiding the excesses of the sub-prime lending prior to the onset of the global recession, their heavy offshore borrowing created a liquidity risk that also rendered them too-big-to-fail, eventually leading to the Government’s funding guarantee. Many American and European banks became insolvent on the asset side of their balance sheet, due to holding dodgy loans and derivatives. But Australia's banks also faced insolvency, except that it was on the liability side of their balance sheet. Again, a greater exploration of these issues is found in The Great Crash of 2008.

Supply-side squeeze?

A common view held by the mainstream Australian press and property 'experts' is that Australia's strong house price growth is due largely to a chronic undersupply of homes. Take, for example, this recent statement from Industry Research and Forecasting firm, BIS Shrapnel:

An undersupply in the Australian property market will force residential property prices up by 30%, according to a BIS Shrapnel chief economist. Frank Gelber gave members at a Real Estate Institute of Victoria lunch a very optimistic forecast for house prices, as reported in the Sydney Morning Herald. "At the end of the day, we haven't got a bubble in our residential market. We're undersupplied, not oversupplied … [House] prices will go up another 30 per cent over the next three years... We’re not over-geared, we’re not overvalued and we’re not oversupplied," he said. ''I can't remember in the last 30 years a time when I have been more comfortable and optimistic about investment in the market.'' Australia’s recovery would not be affected by global economic uncertainty, he added.

Observers from countries that have experienced housing booms and busts would be familiar with this kind of economic delusion. Remember this testimony from Federal Reserve Chairman, Ben Bernanke, in 2005:

House prices have risen by nearly 25 percent over the past two years. Although speculative activity has increased in some areas, at a national level these price increases largely reflect strong economic fundamentals, including robust growth in jobs and incomes, low mortgage rates, steady rates of household formation, and factors that limit the expansion of housing supply in some areas.

How about the classic land (housing) shortage argument previously used to justify sky-high house prices in Japan, the UK, USA, and Ireland: (from Forbes)

Leading up to Japan’s bursting real estate bubble in 1990, we were told that this time was different: house prices simply reflected the growing economic power of Japan Inc. and there was a land shortage in overcrowded Tokyo. Today, Japanese real estate prices are still less than 50% of that peak value reached more than two decades ago. During the U.K. bubble, Brits blamed zoning requirements for creating land shortages until prices crashed to the lowest multiple of income on record in 1997. Similarly, builders in Northern Ireland predicted severe housing shortages if planners did not release more land in 2003. And before America’s bubble burst, there were land shortages across the country, from Florida’s undeveloped west coast to the deserts of Nevada .

So is Australia different to these other nations? Is Australia's housing shortage genuine, justifying the high price tag? Or is it a case of same story, different bubble? To investigate this issue I have first plotted the average number of people per dwelling and the average dwelling size, as measured by the average number of rooms per dwelling:

Click to enlarge:
Click to enlarge

According to the ABS, the average number of people per dwelling has fallen significantly, while the average number of rooms per dwelling has been increasing.

In fact, the number of people per dwelling in Australia has fallen steadily for the past 50 years, from 3.6 in 1960, 2.75 in 1990, 2.62 in 2000, and 2.56 in 2008. So the growth in the number of dwellings has actually outstripped the increase in the population and, therefore, the average number of occupants per dwelling has fallen considerably. Only in the past year or so has the rate of new building fallen behind population growth, as evidenced by a small increase in people per dwelling in 2008. Based on these measures, therefore, there is little evidence of a chronic housing shortage in Australia.

Second, I have plotted long-term house price growth (left panel), population growth (middle panel), and dwelling construction relative to population growth (right panel), in each of Australia's capital cities/states to determine whether there is any obvious link between population growth/dwelling construction and house prices (see here for the full study). For if Australia's house price growth had been driven by fundamental supply/demand imbalances, you would expect price growth to have been far stronger in high population growth/low dwelling construction regions than in low population growth/high dwelling construction regions.

Click to enlarge:
Click to enlarge

As you can see, the macroeconomic data is inconclusive. Whilst tight housing supply goes some of the way to explaining the price out-performance of Brisbane (Queensland) and Perth (Western Australia), Sydney (New South Wales) experienced the lowest house price growth over this period despite having the third lowest level of housing construction relative to population growth. By contrast, both Adelaide (South Australia) and Hobart (Tasmania) have experienced robust house price growth despite both low population growth and high levels of housing construction.

Finally, the below RBA chart shows that house prices have risen equally in Australia's capital cities and regional centres. If housing shortages really could explain Australia's house price growth, we would expect to see city prices climbing much more than in regional areas, since the cities are where the bulk of the population growth has ocurred and where supply-side constraints are more likely to exist.

Click to enlarge:
Click to enlarge

Although the evidence of supply/demand imbalances in the Australian property market is, at best, inconclusive, it is possible that speculators and 'flippers' might have been drawn into Australian housing because of the perceived shortages, thereby fuelling price growth.

In addition, had Australia's urban planning structure been less restrictive and more market-oriented (like the Texan model), it is possible that house prices would not have risen as far, since supply would have adjusted faster to the increased demand from prolificate lending.

Nevertheless, based on the evidence above, it appears that Australia's supply-side situation is not all that different from the bubble markets of Japan in the late 1980s, the U.K. in the 1990s, or Ireland, California, Nevada and Florida more recently.

Bubble trouble?

In my opinion, an Australian house price correction is highly likely in the medium to longer-term. House prices relative to incomes and rents, as well as household debt levels, have simply become too high to be maintained, and all it will take is either a change of sentiment, a deterioration of Australia's economy, another global credit shock, or shifting demographics to bring house prices back into line with incomes and rents. If there is one thing that the global recession has highlighted, it is that asset markets are fluid and can change course abruptly and viciously.

So what are the key risks facing Australia's housing market? Well there are several, some short-term and some longer-term.

First, Australia is heavily leveraged to China, so any slow down in the Chinese economy would likely translate into lower prices for Australia's two largest exports - iron ore and coal - along with other commodites. Should commodity prices fall significantly, it follows that less money will flow into the Australian economy, resulting in lower spending, a contraction in aggregate demand, and higher unemployment. A slowdown of China's economy is highly probable since its two major export destinations - the U.S. and Europe - are facing an extended period of sub-par economic growth from deleveraging as well as aging populations. To date, China has undertaken massive government stimulus to subsidise production and consumption in order to maintain high growth rates. But as highlighted recently by Jim Chanos, such an approach inevitably produces vast amounts of waste and is unsustainable. Futher, there is evidence that China's housing market is a bubble waiting to burst, and has the potential to dramatically slow China's economy.

The second major risk for house prices is that world credit markets freeze, thereby dramatically increasing Australian banks' cost of wholesale funding and/or reducing their ability to raise funds offshore. Recent analyses by investment banks Credit Suisse and Goldman Sachs estimate that the Australian banks will need to raise between $100 billion and $170 billion of term wholesale debt in the 12 months to March 2011, of which about $78 billion would be the refinancing of existing borrowings that mature.

According to Credit Suisse, the refinancing challenge over the next five years amounts to about $420 billion, with CBA and Westpac accounting for about $240 billion of that. The average term of the majors’ debt is only about 2.8 years.

Goldman estimates that if there were 8 per cent loan growth and no growth in deposit funding relative to the status quo, the banks would need to raise $416 billion in five year’s time. If, as is more likely, deposits in the major’s part of the banking system grew by 6 per cent, they would still need to raise $272 billion in five year’s time. That’s additional to the borrowings that will need to be refinanced.

The Australian banks' wholesale funding task will be both difficult and costly since they will be competing with up to $US5 trillion of existing European bank funding that will mature and have to be refinanced by European banks over the next three years, about $US3 trillion of it by the end of 2012. To fund the measures they took in response to the global recession, Australia's banks will also have to compete to raise funds in capital markets against sovereign governments.

Let's also not forget that the Basel Committee on Banking Supervision is looking to add additional layers of counter-cyclical capital buffers to bank capital adequacy requirements in order to mitigate against excessive credit growth and systemic risk.

With all of these factors combined, it will be difficult, if not impossible, for Australia's banks to continue to increase lending for housing. As such, there is little prospect of continued strong house price growth, since continued price appreciation relies on ever-increasing credit growth and household debt levels. Even maintaining current house price levels will be difficult in this new credit-constrained environment.

Even if Australia miraculously manages to dodge the above economic bullets, house prices will still likely come under pressure. This is because for house prices to continue rising, investors and owner-occupiers must continue to believe that capital appreciation will be sufficient to cover the negative income returns from owning residential property. This situation is clearly unsustainable. Once the expectation of continued strong house price growth disappears, households will likely start reducing their borrowings (deleverage) and prices will correct.

Finally, one of the key drivers of Australia's strong house price growth has been the Baby Boomer generation's rampant buying of investment properties to fund their retirement. But with the Baby Boomers soon to enter retirement, it follows that their appetite for investment properties will shrink, thereby removing one of the key demand-drivers of house price growth in Australia. Further, because higher investment yields can currently be earned by placing their funds in bank term deposits and fixed interest than can be earned via rent, it is likely that many Baby Boomers will sell their property investments to fund their retirements. This process of property divestment is likely to accelerate once they realize that there is little prospect of continued high capital appreciation.

The bottom line is that the downside risks to Australian house prices far outweigh the potential upside. Australian housing is simply the wrong asset, at the wrong price, at the wrong time.

Disclosure: No positions