Will the housing soufflé rise twice?

Australia's housing debt has hit the historic highs seen in Japan and the US before their property bubbles burst. Will there be a similar deleveraging trigger here?

One of the advantages of being overseas right now is that it takes less effort to avoid listening to the insipid discourse that passes for political debate during this Australian election. As Kevin Rudd and Tony Abbott compete over who can be more obnoxious to refugees, I find myself pining for the days of sensible and humane policies under Malcolm Fraser. As politicians make themselves the butt of their own unintentional suppository jokes, I pine for someone who could deliver a killer line against his opponent, rather than against himself: Paul Keating.

Whatever one might say about his policies, Keating’s political rhetoric was masterful. My favourite among his many put-downs of opponents was his reference to a pre-Rudd political comeback kid, the flamboyant Andrew Peacock. Peacock beat the dour Howard to become Opposition Leader after Hawke’s victory in 1983, then resigned in office in response to Howard’s white-anting of his leadership (plus la change?), and then came back again after Howard lost the next election. Commenting on Peacock’s chances of winning the election after that, Keating quipped that “a soufflé doesn’t rise twice.”

It’s not a Keating original , but it’s a goodie. And it segues to my topic, a soufflé that clearly is rising again right now: Australian house prices.

Graph for Will the housing soufflé rise twice?

Having peaked in June 2010 (when the ABS index hit 261) after the fuel from Rudd’s First Home Vendors Boost ran out, the index fell by 10.2 per cent to 234.3 in September 2012. It has risen since then by 4.3 per cent to 244.4 in June 2013. This performance has the property lobby bulls and trolls salivating over the vision of a revived investor’s Magic Pudding of ever-rising house prices, while more measured voices worry that another damaging property boom might be upon us (Beware the mother of all housing booms, August 13).

I think they’re both wrong, but it’s also clear that Australian house prices aren’t undertaking the long slow burn down that I expected when I drew the analogy to Japan’s house prices (by saying that Japan’s prices had fallen about 40 per cent over the ten to fifteen years after its bubble economy burst, and I saw no reason that Australia would be different).

Clearly I was wrong here: there are substantial ways in which Australia is different. But they’re not ones that are likely to make the property lobby happy.

First and foremost is that I was wrong in my expectation that Australia (and the OECD in general) would reduce private debt in the same way the Japanese had, and therefore with falling debt would go falling asset prices (since they’d been driven higher by leverage in the first place).

Clearly that hasn’t happened in Australia – or hasn’t happened yet.  But it did happen in Japan after its bubble economy burst, and in the USA after the subprime bubble burst. The differences in debt dynamics go a long way towards explaining why the property markets of these countries have behaved so differently, so that there’s now no way that I’m going to win the Keen-Robertson bet in the way that Rob Burgess interpreted it (a 20 per cent fall 5 years after the bet itself, rather than 5 years after the peak in June 2010, which is my interpretation), though the jury is still out on where house prices will be in June 2015 (see Figure 2).

Graph for Will the housing soufflé rise twice?

So why is Australia different? Mainly because it hasn’t delivered whereas the Japanese and the Americans have.

Until recently, making this point involved some hand-waving, because Japanese data was extremely difficult to locate and collate. But the BIS (Bank of International Settlements: www.bis.org) has recently helped out here immensely by releasing two databases – one on private debt levels and the other on property prices – where the definitions are as consistent across countries as the BIS could make them. This data is very new: the credit data’s existence was announced in June at Malaysia’s central bank conference, and to my knowledge the property data came out after that. It’s going to take a while for me to import it into my own database and analyse it, so this is just the first in a series of posts on both debt and house prices.

These BIS figures have turned upside down some relativities that I thought applied when I had cobbled together data from the USA’s excellent Flow of Funds database, Australia’s good Reserve Bank Statistical Bulletin, and the patchy and badly structured data that I could find on Japan, for two reasons.

Firstly, I used to include financial sector debt in my calculations for the USA, in the erroneous belief that the Flow of Funds data was on debt by non-bank financial corporations to banks. As I noted in an earlier post, that would be the case if I or any other endogenous money advocate had directed the collection of statistics; but it ain’t so, because most economists think private debt doesn’t matter, and they can’t see why banks make any difference to the nature of debt. So statisticians have thrown a grab-bag of debts into the financial bin, and though there is undoubtedly some of the kind of debt that I argue matters in there, it’s just too difficult to tease it apart from the rest.

The BIS – which has the best record of all international economic organisations on the crisis, thanks to its then research director Bill White’s expertise on Minsky’s Financial Instability Hypothesis – avoided this problem by only recording the debts of the household and non-financial corporations.

Secondly, the BIS data – at least on Australia – differs from what is published by the Reserve Bank in its Table D02 (see Figure 2). This must reflect different information that the BIS requires the Reserve Bank to supply, or data that the Reserve puts in other tables being included in the BIS data set, because the BIS – the Central Banker’s Central Bank – gathers its data from its members. 

Graph for Will the housing soufflé rise twice?

There is no trend to the difference in the data, and the correlation coefficient of the two series is very high (over 0.99), but it makes a great difference when comparing Australia’s debt levels to other countries. With my earlier Reserve Bank and Federal Reserve data sources, it appeared that America’s private debt level exceeded Australia’s. Now, on the BIS data, Australia has the dubious distinction of carrying even more private debt that the USA – though Japan remains the all-time champion, at least in this three country comparison (see Figure 3, where the vertical markers “BE” and “BNP” refer to the bursting of Japan’s “Bubble Economy” at the beginning of 1990, and the BNP ushering in the Global Financial Crisis with its shutdown of its Subprime Funds in August 2007). 

Graph for Will the housing soufflé rise twice?

I expect that anyone looking at that chart might think that there’s nothing exceptional about 1990 in terms of Japan’s debt, and that private debt in the USA and Australia kept rising after the BNP decision, so what’s the big deal? The big deal is the turnaround in the rate of growth of debt, which Figure 5 displays. Japan’s bubble economy burst when the meteoric growth in private debt peaked and then fell in 1990, while the BNP decision also marks the peak of debt growth in both Australia and the USA.

Graph for Will the housing soufflé rise twice?

It’s also obvious that credit growth turned negative in both the USA and Japan after their crises – but not in Australia. Credit growth slumped from plus 25 per cent of GDP in Japan at the height of its bubble to minus 15 per cent a decade later, and it spent a full decade in negative territory. America plunged from plus 15 per cent in 2008 to minus 5 per cent in 2010, and it spent two years in negative territory. In contrast, Australia’s credit growth peaked at almost 25 per cent of GDP in 2008 – substantially higher than the USA’s at the time – and then fell to plus 2 per cent in 2010, but it then bounced, rising to a peak at plus 10 per cent in June 2012.

So why did Australia not deliver, and is this a good thing? I had hoped to get on to those topics in this post, but the workload in loading the new data from the BIS got in the way. To be continued.