Danny Yee >> Travelogues >> Oxford Blog >> Books + Ideas
spires from Carfax

UK and Australian housing bubbles

Books + Ideas, , — September 2010

The Economist's global house price comparison suggests that UK residential property is 34% overpriced and Australian property 61% overpriced. Jeremy Grantham looked at 34 historical asset bubbles, of which 2 remain unburst - the Australian and UK housing markets. In the Demographia metropolitan market affordability rankings (PDF), Australian cities take six of the bottom ten places, with median multiples from 7.4 to 9.1, while London and significant numbers of other UK areas are over 6.

On any formal metrics - the Economist used price-to-rent ratios, Grantham used deviations from long-term trends in price-to-income ratios, Demographia uses median price to median income ratios - it's clear that Australian and UK house prices are at or near all time highs. Reversion to historical values for these metrics is not a certainty, but given comparisons both with other countries and historically within the UK and Australia (e.g. the 1890s), there has to be a good chance that prices will return, at some point in the future, to historically normal levels. (Grantham: "it would be the first time in history that such a bubble has not broken".)

House prices are now driven largely by the availability of credit, so they are most sensitive to interest rates and required loan-to-valuation ratios. And the ability of the private sector to increase debt levels seems limited. The following graph (taken from Steve Keen) provides some historical context here.

I find none of the counterarguments to the likelihood of a crash or at least a correction at all convincing, especially given the events of the last few years and a comparative perspective. Many places have worse housing shortages (and Australia's ratio of people to houses has actually been dropping, not increasing). Long term demographic trends such as the shift to two-income households are also present in Spain and Ireland and the United States. Finance and tax advantages for housing are also widespread - in the US, mortgages are typically fixed for thirty or forty years, and mortgage payments by owner-occupiers are tax deductible.

Which will go first?

Judging by the areas I've looked at first hand, rental yields in Oxford are definitely higher than in Sydney - gross yields up around 5% instead of well under 4%. So the Economist numbers seem right to me: the Australian bubble is bigger.

However the Australian economy is steaming along just fine, while the UK has higher unemployment and is braced for massive cuts to the public sector, with accompanying redundancies. The Australian government also has less debt, giving it more room to bail out the private sector in the event of further financial shocks. And the devaluation of the pound has already lowered effective house prices for foreign buyers (most citizens of the EU have the right to live in the UK and buy property here) and locals with foreign assets.

It's hard to see anything rapid happening in Australia, barring sudden shocks from China. On the other hand, the banks' cost of funding is likely to increase, even without RBA interest rate moves, and they are unlikely to go back to 95% loans, making it hard to see how significant further house price rises can happen. But even flat house prices for any extended period of time will make negatively-geared investors (about two thirds of them!) unhappy, so the ponzi part of the bubble has to unwind eventually...

As Keynes is supposed to have said, "markets can remain irrational longer than you can remain solvent", so I'm not about to short the Australian banks (or even go short CBA and WBC and long ANZ and NAB, which would be a safer approach). I am, however, keeping my hands out of the FIRE.


  1. And Goldman Sach's response to the Economist estimates that the Australian property market is only between 24-35% overvalued. And that the trends of paying down debt and a slowing in the increase in house prices during the recession indicate it is not a speculative bubble, and if those trends continue with rising interest rates the situation will improve. There are other trends like the fact that the percentage of migrants setting in Sydney has fallen from 50% to 30% over the last decade because of a shortage of housing amongst other reasons.

    They also noted that it would be a good idea if Swan stopped using the RBA to attack banks for raising interest rates to reflect increases in credit costs for short-term political reasons.

    The Economist and other UK analysis of the rental markets also don't understand negative gearing in Australia and its effects on increasing property prices and driving down rental return figures. I also notice that they use long-term ABS average rental figures which are very low. While rent increases stopped during the recession the Dept. of Housing found that they had resumed by the end of 2009 and that in Sydney there was a 2.8-6.5% increase in rents for two-bedroom apartments in the last quarter of 2009. The cost of renting a two bedroom 1960s apartment round here has risen to $400/wk or $20,800 pa (though so far the Government's tightening of "skilled" migration residency rackets has only affected vocational eduction and not the major universities).

    Most negative gearers will not panic if property prices stop rising, or even fall a bit. They take their tax subsidies now and plan on taking their capital gain in 10 or 20 years time. Those already in the market are currently sitting on capital gains of 165% of the last 10 years and 289% over the last 20 years (and that gain is worth twice as much as on cash savings due to tax law). It is only recent entrants who don't have the income or assets to cover their position who will be burnt by a slowing or slight fall.

    The arguments about long-term per capita housing are also misleading and largely irrelevant due to social changes. Everybody knows that shortage is partly caused by smaller families buying or renting much larger houses using two incomes (or one income plus welfare). And that there is a real shortfall of 250,000 dwellings on demand and new home construction in NSW is very low. That demand is not totally inelastic but unless families start renting out their home theatre rooms, in far flung estates, for share accommodation it isn't likely to change.

    In the US debt levels were much higher than Australia, cash interest rates at approximately zero in real terms for over a decade and most mortgages were non-recourse and fixed term with a 2% discount for the first two years. There are people there choosing not to pay their mortgage and investing in savings because that is in their interest and on the downside they get to live rent free for upto 2 years until the mortgager forecloses on them. Different situation.

    It is also worth noting that despite being in a much worse economic position with regards to economic growth, austerity measures and national debt housing prices in the in UK have dropped nearly 5% toward more sustainable levels without any bubble imploding so far.

    Comment by David Watford — September 2010
  2. Goldman Sachs may not call it a speculative bubble, but a 35% revaluation would be more than enough for most people to consider it a crash. I wonder if Fitch will stress test the banks against that kind of scenario?

    Negative gearing is not specific to Australia - in the US owner-occupier mortgage payments are tax-deductible! - and in any event doesn't change the fundamentals of housing as an investment. As all the timber MIS investors have found out, a bad tax-deductible investment is still a bad investment.

    Most investors probably will sit tight even if underwater, but prices are decided at the margins.

    People were talking about a housing shortage in California in 2007, too. There's always a housing shortage, because everyone wants to live in a four bedroom house on a quarter acre block. As in California, that demand will prove extremely elastic once it turns out that house is not a money-spinner as well as a nice place to live.

    Ireland has full-recourse mortgages. See how well that's stopped a crash there! None of the claims that Australia is unique hold up to even cursory examination. Everything may be running fine at the moment, but at current debt levels even a small hiccup in the economy could cause big problems. It's not impossible, but I think it would take above-trend growth and some tightrope walking to unwind the ponzi elements of the last decade of house price appreciation without a crash.

    Comment by danny — September 2010
  3. Looks like Fitch is stress testing the banks against a 40% fall in house prices. That will be an interesting exercise, but I can't see that scenario eventuating in the near future - the people shorting banks may be a little early.

    Comment by danny — September 2010
  4. Nice article. Couldn't agree more. I have written a similar detailed examination of the Australian housing bubble on my blog:


    Comment by Leith — September 2010
  5. A 35% collapse in property prices overnight is a crash.
    A 35% unwinding over a decade by not increasing as much as GDP and paying down debt is not.

    Despite 20 years of high house prices nearly 60% of owners have bought or are paying off a house to live in, not to speculate on.
    Most landlords own properties for long term investments. And as long as your working age population grows your GDP will grow and land will increase in value in the long term, even if it goes down and up at times.

    It is only flippers trying to make a quick million who are a problem. And they are only a small number and the real problem is that if property values drop sharply they may be forced to sell at large losses further driving down property prices and causing an overshoot in the reduction in property values.

    Negative gearing is a much greater benefit in Australia than anywhere else because it can be claimed against other income. It does raise rental property capital values and it means that real rental returns are higher than the economists figures. No it will not save you from speculating on capital gains if the prices falls.

    And you keep comparing Australia to the US and Ireland. Both had household debt levels twice that of Australia. Both had very low interest rates which Australia didn't. Both had severe recessions, in the case of the Ireland a 20% contraction in GDP because there economy was driven by tax subsidies, debt and EU money.

    Despite interest rate hikes, a recession (fairly mild), and further rate rises to come Australia's foreclosure rate was 0.17%. Most households are paying 30% of income on mortgages. Which they have been for a couple of decades and is similar to what they would pay in rent. The argument that after 20 years that is unsustainable because people paid less several generations ago when circumstances were completely different ago is nonsense.

    And California was building massive numbers of new houses, they people ended up never being able to get the credit to buy and move into. In NSW new housing constructing is down 16% this year.

    Australian property prices fell during the recession. They grew from the end of 2009 but that was just because people were rushing to catch stimulus measures like first home buyer discounts. When they ran out in June prices stopped rising and fell 1.2% in the September quarter. Mortgage holders have been paying down debt in anticipation of further interest rate rises.

    Yes if Australia's GDP suddenly collapses by 20% there will problems, but you aren't ever going to escape problems if that happens. But at the moment the situation is sustainable and improving slowly.

    Comment by David Watford — October 2010
  6. It would clearly be a good outcome if Australian housing prices can drop 35% over the next decade, steadily and without any crash. But achieving that - possibly by sustained high inflation rates - seems like a near-impossible tightrope-walk to me. At some point, with flat or falling prices, investors are going to realise that negative-gearing means losing money.

    The demographic and comparative arguments are too much like special pleading. You can see this from their mostly being invertible. So Australia having full-recourse mortgages may make people more reluctant to give up paying their mortgages and thus make a crash less likely. But full-recourse mortgages should also make people warier about taking on large mortgages, and hence help to keep prices lower. And the demographic shift to smaller households has increased demand for housing. But when that shift stops or slows, as it has to, that demand will go away (and the further that shift goes the more room there is for movement in the other direction). On the subject of demographics, this BIS paper http://www.bis.org/publ/work318.htm suggests the same shift that helped propel house prices over the last forty years will provide negative pressure of the same magnitude over the next forty years, OECD-wide (see Graph 3 for the key result).

    Australian house price to income multiples in major metropolitan areas are as bad, or worse, than the multiples were in California or Dublin at their peaks. And rental yields are terrible. For these numbers not to be significant, the Australian housing market would have to be qualitatively different to every historical housing market, ever. (And in some ways, to every other asset market, ever.) I just don't buy that. (And if you're accepting we're likely to see a 30% fall in prices over some time frame then maybe you're not really disagreeing about this.)

    Comment by danny — October 2010
  7. I agree with you than any kind of rapid crash is unlikely while the economy is growing and unemployment is low, though it's not impossible that a housing market crash could lead the economy into recession rather than the other way around. Here's another analysis http://blog.lvrg.org.au/2010/10/40-years-of-housing-bubbles-and.html They use price to per-capita GDP, which suggests less of a bubble than CPI-adjusted "real" prices.

    Comment by danny — October 2010
  8. Every housing market today is complete different to housing markets one, two, three of four generations ago.

    And again that blog over states the overvaluing by using the economist's rental yield figures. Those figures use national average rental figures and omit negative gearing tax savings and also that the capital rise in dwellings is two to three times greater than average is in the inner city areas of Sydney and other capital cities.

    Something that grows for 30 to 40 years then goes down when you run into 1 in 50 year recession or depression which kills demande is not a bubble. That is just the economic cycle.

    A bubble is a short term rise driven by people chasing capital gains, commonly fuelled by cheap debt or government stimulus, that collapses because the price stops going up.

    Ireland did not suffer from a housing bubble, it suffered from a mass of economic problems, most of them externally driven. Housing is the least of their problems.

    The US did suffer from a housing bubble, though more importantly it suffered from a bubble in securitised mortgages and then derivatives on them. A bubble large enough to send only the US economy but much of developed world into recession.

    All totally different to what has happened before, every bubble and recession is unique because since you know what happened before you always have to come up with novel variation to create a new one.

    Demand in Sydney is currently inelastic not elastic. The inner city is where the jobs, services, education, transport and best climate is and construction of new dwellings are down. Yes it can become elastic but it is going to take a lot. California's demand didn't become elastic by people redistributing themselves over the housing, they left new estates empty and moved to Ohio, Texas and Georgia. Where are you going to go if you leave Sydney?

    Australias' economy handled the last two recessions, which were mild, by cutting hours and not jobs. Sure there will a recession some time in the future that could be much worse but no on the immediate horizon. There is no interest spike from the RBA because they are already at or above the long term average.

    The main risk of a large rise in interest rates comes from the fact that 50% of mortgage credit comes from OS, but then Australia is the best place to loan money at the moment.

    Comment by David Watford — October 2010
  9. And Recourse loans have made a difference in Australia.
    Homeownership has fallen from 77% to under 60% over the last 30 years.

    Comment by David Watford — October 2010
  10. Ireland clearly had a housing bubble. You won't find many people denying that, at least not after the event! And there were other problems as well, but it's the losses on property loans that pushed Anglo Irish and Allied Irish to the wall.

    Property investors in Australia are clearly chasing capital gains, as evidenced by the fact that two thirds of them are negatively geared, with running yields lower than their holding costs. The tax deduction just reduces the amount they are losing.

    Where do you find figures for home ownership falling from 77% to 60% over the last 30 years? The Parliamentary Library report at http://www.aph.gov.au/library/pubs/rp/2008-09/09rp21.pdf gives ABS figures of 68% for 1976 and 70% for 2006, and it's never been anywhere near as high as 77%. (It also seems most unlikely to have dropped 10% in the last four years, though we'll have to wait on the next census for the next number in that series.)

    Comment by danny — October 2010
  11. Ireland did not have have a housing bubble. You can't label every assets that falls a bubble after the fact.

    The EU started giving Ireland development money, which Ireland used to fund tax reductions for foreign companies. With low land prices, low labour costs and an educated English speaking population the low taxes made Ireland the place for Asian and US companies seeking to get into Europe to locate.

    The Irish economy grew at nearly 10% pa for 10 years. Of course this fuelled consumption, assets and property booms. However since the companies were multinationals they transfer priced the profits out of Ireland and there was a 25% gap between GNP and GDP and the tax discounts meant they weren't funding Government's increasing revenue causing government debt. As land, labour and taxes rose they started moving on to East Europe, the next big thing.

    Being part of the Euro, so unable to control interest rates or the currency, and with increasing Government debt the Government tried to make austerity measures. This further reduced GDP while consumption, assets and property prices rises continued to be funded by more private debt. At which point the developed world economy when into recession and credit dried up.

    The Irish banks were in trouble because debt was being used to fund consumption, commercial property and because they were heavily involved in global securitised mortgages and derivatives trades as well (just like just like Iceland and most of the other European Banks).

    Their domestic housing problem is only one part of that. And more caused by the economy growing by 150% over a decade, then collapsing 20% overnight than by speculation.

    Of course in trying to reduce debt the Government's austerity measure reduced revenues further, and because of the global crisis that got worse and they have had to bail out all their banks, and credit became more expensive, especially as they are the 'I' in 'PIGS". So there GDP has contracted 20% and their debt as increased quite a lot.

    Comment by David Watford — October 2010
  12. My figures for home ownership were from a paper from a property reasearch group quoted in the SMH this month, that claimed home ownership had reduced from a peak of 77% to 60% over decades, and predicting strong growth in rents as a result. There was another article in the paper claiming that after the recession, interest rise and shortage of credit it was currently as low as 56%.

    Housing markets commonly unwind slowly rather than collapsing, in the absence of any economic crisis. The primary reason that housing exists is to house people. And I say 60% and you say 70% of the market is made up of people owing or paying off a home to live in. 5% of the market is public or community housing, then there is the aged care sector. Landlords only make up the rest. If prices go down owners and long term investors have an incentive not to sell into a falling market, crystalise a paper loss and cause a positive feed back causing prices to undershoot.

    It is only highly geared speculators,and people who become unemployed who can't service their debt who sell into a falling market.

    Most people who negatively gear property will only loose profit if the market stagnates, not be forced into causing a price spiral. The idea that all negative gearers are mad people with all their assets gambled on capital rises in the property market, like Wall Street bankers, is nonsense.

    My sister has a negative gear property but is in no danger of losing money on it. She moved in with her boyfriend and got married and negatively geared her house. She has fixed the roof, will put new carpets in. She is now looking to sell and make quite a bit of money on having done so. She has plenty of equity in the house so if prices collapsed tomorrow she would simply make less capital gain. Sure that means she may have been better of selling the house a few years ago and investing in cash or paying down the mortgage on their current house, but she will still have made some money out of it.

    I have another friend who has been negatively gearing apartments for 25 years. He has recently been made redundant so has sold a few and diversified to cope with semi-retirement. Even a large collapse in property prices would just leave him making a reduced profit on the apartments he still owns.

    Even in the US market if you bought before 2004 on average you property today is worth what is was then and you will likely make a capital gain in the future. It is only those who bought during the rapid increase in prices as the bubble inflated between 2005 and 2008 who are in trouble. The later they bought the more trouble they are in.

    In the next 20 years the Australian population is going to go from 22m to 30m due to increases in population even with an extremely low migration rate. At current rates it will hit $35m.

    Comment by David Watford — October 2010
  13. Thinking about it the 56% figure was probably from the Herald's special feature on Sydney property prices last week, so is probably a Sydney not a national one. Sydney is the least affordable market in the country.

    Comment by David Watford — October 2010
  14. I'll take the ABS over the SMH for statistics, methinks!

    Denying that Ireland had a housing bubble seems very odd. There are a small number of economists who deny that asset bubbles are possible at all - basically adherents of extreme versions of the EMH - but that seems untenable outside pure theory.

    If you accept that asset bubbles and crashes exist at all, then the Irish residential property market of the last decade is surely a clear-cut example, with a rapid fall of 40% or more and no sign of any recovery. And Irish and UK newspapers - and the Wall Street Journal, the NYT, etc. - all use terminology like "housing bubble", "property crash", etc. quite regularly in talking about Ireland, without any qualification.

    The UK, on the other hand, has not had a housing crash (not since 1989-94, anyway, when real prices fell about 35%).

    Comment by danny — October 2010
  15. There are bubbles, just less than the media and doom mongers claim. Tulips for example were not a bubble because they weren't even an asset but a consumable good. They were a fad. Wealthy houses in Holland and England took to using Deft standing vases of tulips as elaborate hall and table decorations, then they switched to things like porcelain figures and marzipan sculptures. The very high high prices quoted by economists and the media for a bubble where the top prices for a few rare types of tulip and not a median price.

    Ireland was not a property bubble it was a severe recession. Property prices are never going to the same in economic boom as in a severe recession because demand changes massively.

    They aren't SMH statistics they are property researcher reports. And as for reliability of the media your prime data was a report by the Economist on a foreign market and some attention-seeking bloggers.

    They claim that Australian banks would be in trouble if housing prices fell. Fitch's preliminary report this week found that a sudden 40% drop in prices would cause $9b in losses for the banks, which wouldn't even wipe out half the annual profit of the big four.

    Looking at the ABS debt to disposable income if look at just the mortgages house prices rising from 3 times median disposable income in the early 90s to 4 times median disposable income now. Which has been done an increase in the size or mortgages from deregulation.


    A historical argument that things today return to exactly as they were when strict regulations restricted credit, the size of mortgages and repayments schedules is obviously flawed.

    That deregulation has also resulted in longer mortgages, lower fees, lower interest margins and flexible mortgages where people can adjust the rate they pay it back depending on their circumstances so repayments are not that much higher. Nor at they much more than rent.

    Comment by David Watford — October 2010
  16. You can define things however you like, but the tulip mania is one of the archetypal examples of a speculative bubble (wheat and oil are also perishable consumables) and everyone else on the planet thinks the Irish property market was a bubble. It takes more than a severe recession to drop house prices by 40%.

    Greater availability of credit has indeed pushed house prices higher. However these changes can't continue indefinitely - since we're unlikely to go past 40 year interest-only mortgages at 110% LVR - and are quite capable of going the other way. As indeed they are - slowly as yet in Australia, but rapidly in the US and elsewhere.

    The Irvine column seems weak to me. The only evidence she adduces against Grantham, the Economist, et al. is the shift to dual income households - which has occurred in the United States and Ireland and Spain as well. Oh, and Sydney is apparently surrounded by mountains - except that the flat area between the Hawkesbury/Nepean and the sea is bigger than the Greater London Urban Area, which has more than twice Sydney's population (and that leaves out the Central Coast, Blue Mountains, etc).

    Comment by danny — October 2010
  17. There was clearly a bubble in the Irish property market.

    The head of the Central bank of Ireland called it a bubble several times in a speech to the British - Irish Parliamentary Committee.


    Comment by Ramal Gnanasekaran — November 2010
  18. Here's a nice graph of Melbourne house prices. From the 1890 peak it took 110 years for real house prices to get back to the same level. Could it be 2120 before they get back to the 2010 peak?

    Comment by danny — April 2012

TrackBack URI

Leave a comment

Books + Ideas << Oxford Blog << Travelogues << Danny Yee