It's the housing crash, stoopid!

Gavin R. Putland writes:

Even if your theory is sound, economic forecasting is an uncertain business. It is a near-certainty that a slump in property prices is soon followed by a contraction in GDP; but when asked how soon is “soon”, one cannot be so confident. Hence the probabilistic tone of my remarks in today's edition of Crikey: “I'm expecting tomorrow's GDP figure for the March quarter to be worse than the median forecast of mainstream economists. Probably much worse.” (Emphasis added!)

Confidence improves considerably when one allows a bit of wiggle-room about the timing — as in September 2003, when I predicted the GFC.

Actually that was rather a lot of wiggle room. A better example came in November last year, when I predicted an acceleration of the price-slump in the Australian housing market:

The average Australian housing investor is negatively geared. That is, the interest cost incurred by the average investor is not justified by rental income alone, but depends on capital gains as well. The “average” is taken over all investors, including those who own their properties outright and/or bought their properties for far less than their current values. Anyone contemplating paying current prices will need faster capital gains than the average investor. Unfortunately for the prospective sellers, and especially for recent buyers, those capital gains have stopped...

One possible explanation... is that after so many bubble-bursts in so many countries, property owners have decided to get out while the getting is good. That would be consistent with the sudden talk of a “glut”...

But whatever the explanation, prospective buyers who observe the peaking or “flat-lining” of prices will conclude that this is not the time to buy. I would expect that realization to show up as a renewed slump in lending, for both owner-occupation and investment, in the near future. Then I would expect the usual sequence — a slump in sales followed by a slump in prices — to reassert itself with a vengeance.

It turns out that in Q4 of 2010 — the same quarter in which I wrote those words — residential land sales fell to a 10-year low nationwide and a 16-year low in Melbourne. Of course the spruikers tried to explain the low turnover in terms of land shortages, supporting their claims by referring to price rises since a year earlier. But when the market is overpriced, falling turnover is more likely to indicate falling demand than falling supply.

The “renewed slump in lending” soon materialized. In the March quarter, as I noted on May 16, lending to buy established homes for owner-occupation fell to its lowest level (relative to GDP) since Q4 of 2000, while lending to individuals to buy investment homes (again relative to GDP) was the lowest since Q2 of 2001.

Indeed I needn't have taken the trouble of scaling the loan values to GDP. As the media duly reported, the seasonally adjusted number of loans for owner-occupation fell to a 10-year low in February and repeated the performance in March. In the three months to February, loans for the purchase of newly erected dwellings fell almost 36% — the biggest 3-month fall in records dating back 32 years. The corresponding fall in sales is evident in the monthly sales graph from RP Data.

Of course RP Data is best known for its Home Value Index, which peaked in May 2010 and fell by 2.1% (seasonally adjusted) in the three months to March 2011 — the biggest three-month fall since the beginning of the index series in June 1999. The same index declined another 0.3% (seasonally adjusted) in April. Meanwhile the rival Residex index (which is not seasonally adjusted) declined in every capital city in April. The previous two occasions on which that happened were June 2008 and June 1990. Both were followed by contractions in GDP within six months.

The decline in sales and prices has coincided with a blowout in stock on the market. According to SQM Research, the stock of housing for sale rose 68% in the year to April.

There has also been a blowout in mortgage delinquencies. In the March quarter, according to Fitch ratings, the percentages of prime and low-doc residential mortgage-backed securities in arrears by 30 days or more exceeded the records set during the GFC. But this is not seen as a problem, because the overall level of arrears is considered low.

In view of the decline in lending, sales and prices and the blowout in unsold stock, and in view of what happened around the time of the last peak in arrears, I submit that the latest spike in arrears should be seen as a lagging indicator of a deeper problem: that too many people have borrowed too much to gamble on rising home prices.