Monday, April 11, 2011
TMM probably should have realised that by broaching the topic of Australian rates, the composition of household wealth and thus housing we were walking around the caldera of a very active volcano of debate in both markets and policy. You see, some people like Jeremy Grantham have made the point that Australia is a housing bubble because its growth is so far away from trend, others, admittedly with some skin in the residential property game like Chris Joye at Rismark in Australia have staunchly defended Australian residential real estate prices as more business as usual in affordability terms and the rise in values as being due to secular declines in inflation and therefore rates. Steven Keen and Chris Joye have even debated here. So, what does TMM think?TMM think that rather than throwing peanut shells from the sidelines that its worth taking a step back and ask what a housing bubble is anyway. As with everything, there is demand and supply:
- Demand: Real wages, population growth, employment growth, CREDIT, tastes and preferences, cost of capital.
- Supply: Costs of construction, real geographical constraints, supply side constraints like zoning, short term elasticity of supply.
- Market Microstructure (how 1 and 2 sort it out).
You would be inclined to think the whole world has decided that any real estate market comes down to one variable and one variable only, the one in BOLD above - which is a gross oversimplification that allows journalists to lazily fill column space without doing any real work.
TMM thinks we covered the demand side in Australia pretty well (real wages, employment, rates in the previous piece) and Joye makes a good case that given these ratios have been stable since the late 90s as people's tastes and preferences adjusted to new, lower long term inflation expectations that there isn't anything untoward in house prices' rise in the 2000s.
On the supply side there has not been much work done on the costs of zoning in Australia aside from stuff like this but it appears that the costs are mind-blowingly high. More work needs to be done on this by the RBA and their ilk on how the geography of Sydney and its ring of national parks affects housing but its an article of faith that this is a big deal for affordability. Ed Glaeser covers this well in the US and it is worth noting that those markets that are most constrained like New York have bounced hardest coming out of the crisis, Phoenix and Las Vegas, less so.
Perhaps the "tell" of the housing bubble in the US was that markets with vastly different fundamentals started to perform similarly, in that subprime defaults were mere symptoms of the underlying credit bubble that had manifested itself in increasing leverage across all income, demographic and geographical groups in the US. That does not seem to be the case in Australia right now. As far as the distribution of leverage goes, many bears point to the First Time Buyer Credit leading to a ramp in prices in the aftermath of the GFC. TMM are sceptical on this point given that, unless they misunderstand the ABS's numbers, that only around 220k first time buyers took advantage of this perk. In terms of the effect of such a group moving en masse into delinquency, the numbers are just not large enough to produce a meaningful spill over into the rest of the economy (220k * 300k average first time buyer price ~AUD 66bn). As TMM noted on Friday, the Household Net Worth position in Australia does not show any particular sign of overleveraging, sitting at around 600% of disposable income, relative to liabilities of around 140%. A point often made is that using average wealth and income levels masks the underlying skewed distribution, and is one that TMM are always very aware of. However, in Australia, they do not necessarily believe that there is much wrong with this level of abstraction given that Australia's Gini coefficient sits at just 30.5 (25th lowest in the World), vs. 45 for the US (95th in the World, and below that of the Ivory Coast!).
In terms of geographical distribution, Perth and Brisbane are ripping while Sydney has been flat lining for a long time. To TMM that says a lot about what they pay wheel loader drivers vs what they pay lawyers today and indicates that forces aside from credit are at work here. Note that in these regions there was pretty much no one living in many of these areas.
On market microstructure, the picture is reasonably positive. Most houses are sold via auction in Australia which can induce bidders to bid up but also ensures people know where the next best print is - its hard for a market to go completely illiquid when they know where the next bid is which was a serious problem in some areas of the US where housing seemed more like selling CDO tranches in early 08 than anything else.
Then there is the issue of how that market micro-structure can or will handle "weak hands" i.e. - over-levered players who get stopped out and put a lot of supply on the market as was the case with US subprime... or any agricultural futures expiry this year. As far as TMM can tell, those people have largely been blown out already: looking at the hands-down crappiest securitisation in Australia in 2007 (Mobius NCM-04) losses have already been taken and those loans that are still performing are at much lower LTVs. No one prints this stuff anymore and if deals like this haven't taken down the market they are not likely to anytime soon at LTVs <60%.
Finally, is there any evidence of return-chasing, that other key ingredient of a Bubble? As regular readers will recall, TMM have a particular fondness for considering whether a certain market's returns are auto-correlated as this provides a statistical measure as to whether current returns are explained by past returns (i.e. - did punters see that prices had gone up and decide to buy more?). To that end, TMM have compared the 8-quarter rolling 6m auto-correlation of year-on-year house price growth in both the US (see chart below, red line) and Australia (blue line). In the US, it appears that the bubble began to form in the mid/late-1990s, crescendo-ing into 2005 before crashing spectacularly - the strong auto-correlation from 2007-2010 being a function of house price declines leading to even more declines (leverage working in the other direction). But in Australia, aside from a brief period of positive auto-correlation between 2001-04 at low levels relative to the US, something arguably related to the Zoning issues TMM discussed above, it is not particularly obvious that there was any particular degree of return chasing in the Aussie housing market. TMM would also note that Australian house prices fell/went sideways from 2003-05 and similarly from 2007-9: house price falls are not unusual and therefore embedded into household expectations, unlike in the US where house prices had not fallen on a nationwide basis since the Great Depression.
So, are TMM Aussie housing bulls? Hardly. TMM are less concerned about some inherent instability and fragility in the housing market than they are about broad instability and fragility in the Australian economy. While LTVs may be low going back to those Perth wage changes it is quite clear that Australia has an awful lot riding on the mining boom and implicitly Chinese fixed asset investment. Seeing how much of Australia's total investment is in mining and how average wages in suburbs in Perth have consistently outstripped GDP for some time running, we are also concerned about how a real shock (namely, China slowing substantially) would impact real incomes and in turn housing. Australians have a lot of wealth locked up in housing (around 440% of disposable income) and the tax system encourages people to be long property. It is possible that a real shock could lead to a mother of an unwind but once again, that is a shock from an external source, not some inherent criticality in Aussie housing itself.
TMM can't help but feel that Australian housing is riding the Dragon and little else. Then again, just about everything else Australia is, including the AUD. Looking for the next subprime debacle and opportunity to short billions of mortgages for <100bps per year and watch them go to 20c is not something that is going to work in Australia before a couple of other interesting things happen - like copper going to $4000/ton and the AUD/USD revisiting .70 or worse. Put simply, there are easier trades.