Tuesday, February 12, 2008

Housing bubble: dynamics of a bust





The first figure is from today's WSJ and incorporates data from Q4 2007. The second figure appeared in the Economist some time ago and was discussed previously on this blog. Does anyone care to predict the future for bubble states like California, Florida and Arizona using the Japanese data as a guide?

Lower interest rates will not re-inflate the housing bubble (although they may affect the rate at which it deflates; note the BOJ dropped real interest rates below zero in the wake of their bust). People understand now, as they did not just a few years ago, that home prices can go down. This change in ape psychology (try putting that in your macro model!) makes all the difference.

Below is historical data compiled by Yale economist Robert Shiller showing that home prices have not on average provided attractive real returns (right hand axis is inflation adjusted returns for same house sales over time; previously discussed here -- the real rate of return was 0.4% between 1890 and 2004). This is yet another example in which market participants (home buyers) made decisions based on faulty assumptions that might have been easily corrected by a modest amount of research. So much for efficient markets!



Here's some detailed data from Case-Shiller and OFHEO indices (also from WSJ; note OFHEO only tracks conforming mortgages so has less sensitivity to the high end of the market):



Finally, it is worth noting that the subprime mortgage meltdown is merely a symptom of the real estate bubble. If home prices continue to fall we will see (as we are already beginning to) higher default rates in so-called "prime" as well as subprime mortgages.

WSJ: ...I assumed, for the sake of calculations, that California prices fell 8% last quarter from the third quarter, a huge number by historic measures but not out of line with Zillow's data. For Florida and Arizona I assumed declines of 5% and 5.5%. You could use other, more modest estimates for the recent declines: They won't change the outcomes much. I also assumed personal incomes in these states rose in line with recent and historic averages."

The results? In all three markets, the prices are well off their peaks when compared to incomes. But they remain far above historic averages.

Median prices in California peaked in 2006 at 13.3 times per capita incomes. Hard to believe, but true. They may be down now to about 11.1 times.

But that's still way above the ground. Throughout most of the 80s and 90s they ranged between six and seven times incomes.

Just to get down to seven times incomes, prices would have to fall 37% tomorrow.

Those who bought at the peak of the cycle may be pinning their hopes instead on "incomes catching up" instead. But they had better be patient. Even if house prices stayed exactly where they are, it would take around 10 years for rising incomes to bring the ratios back into any sort of alignment.

2 comments:

Anonymous said...

Actually that market is quite efficient :)

30Yr bond price change with 3% yield drop grows ~50%... Not mentionning yield drop from 18% in 70's...
Equivalently, psychologically, ppl care more about their "monthly nut"
and less about TOTAL debt. Especially in the rising price environment. So if one can afford bigger, better house for the same $1000, one moves there emptying his house for a guy down the ladder... Everybody efficiently moves up the pyramid till it flips over :)

And of course some places got multiplier effect from NINJA loans -- ppl are not that stupid like some wish to portray them, if you get an option of upside almost for free, why not utilize it?

On this topic see
http://www.youtube.com/watch?v=l3kaK2_CrEw

Anonymous said...

When did the rapid rise in housing begin? Circa 1997 when we had a change in tax treatment of housing. Is this a coincidence? I think not. We altered a market. Why did we relax sub prime lending? We had a congressional mandate.

Yes lenders have played fast and loose but SO HAVE BORROWERS.

Blog Archive

Labels