Robert Shiller has been predicting
a housing-market crash for as long as anybody, which is why his piece in the
Wall Street Journal today
is so interesting: it seems he’s not predicting a crash any more! Here’s the
final paragraph of his piece in the same space on August 30:
Unfortunately, there is significant risk of a very bad period, with slow
sales, slim commissions, falling prices, rising default and foreclosures,
serious trouble in financial markets, and a possible recession sooner than
most of us expected. Deterioration in that intangible housing market psychology
is the most uncertain factor in the outlook today. Listen hard and watch out.
And here’s his conclusion today:
We are left with a deeply uncertain situation, but one in which it would
seem that a sequence of price declines continuing for many years has some
substantial probability of happening. Traditional finance theory has trouble
reconciling even a semi-predictable sequence of price declines with basic
notions of market efficiency. The situation we are facing is a reminder of
the glaring inefficiencies and incompleteness of existing markets for residential
real estate, and may be regarded as evidence that institutional changes will
be coming in future years to fundamentally change the nature of these markets.
There’s the move from "significant risk of a very bad period" to
"some substantial probability" of "a sequence of price declines".
But there’s also a more philosophical shift: the first piece was often reprinted with a graph
which was clearly designed to show that the housing boom looked just like the
dot-com stock-market boom, and that the housing bust would not be dissimilar.
Now, Shiller’s saying that there’s nothing predictable about housing prices
at all, and that anything can happen:
With the failure of anyone really to predict today’s high home prices, one
may well conclude that no one can predict today whether a home-price bust
is coming, or whether the housing market will land softly, or even is poised
to resume its upward climb. That may be the right conclusion about our ability
to forecast the markets…
We haven’t seen really long and significant strings of price decreases in
the U.S. since the first half of the 20th century. The consecutive-year home-price
declines in the major U.S. cities from 1990-93 amounted to a total drop of
only 8% from peak to trough. More recently, we have seen sharp reversals of
sudden price drops. San Francisco home prices dropped 7% between 2001 and
2002, and then resumed a strong upward climb. London home prices dropped 5%
between 2004 and 2005 and then resumed an upward climb…
Of course, Shiller balances all his upside scenarios with downside scenarios.
But my point is that that balance was nowhere visible in August, when he was
nothing but downside scenarios. Why the change, I wonder?
I found it interesting that Shiller would take Harry Dent to task for not predicting the housing boom, but keep his own housing forecasts hidden from view. Is he a purveyor of truth or of the double standard of self-promotion?
Interesting that Shiller should cite the mild price drop in SF from 2001-2002. I expect history to rhyme (if not repeat) here on the SF Peninsula. But that’s because buildable land is near zero here, so there’s no oversupply.
East of SF, however, large new tracts are still being completed, and prices have already fallen more than 10%. Even the National Association of Home Builders admits to serious oversupply:
The biggest problem the housing market faces is “a seriously large inventory situation,” said David Seiders, chief economist for the National Association of Home Builders, which is hosting the International Builders Show here this week. Seiders said the housing boom in 2004 and 2005 produced at least 400,000 more housing units than demand could support, and builders are having to push hard to move those homes off the market.
http://www.marketwatch.com/news/story/housing-still-falling-midyear-bottom/story.aspx?guid=%7BD3228219%2D2D28%2D4B44%2D8796%2DE63E9AD525ED%7D
Even if the above estimate of 400,000 units is not understated, that’s still a lot of houses! I wonder if there are enough renters in the overbuilt areas to occupy these units. If not, how low would prices have to go to induce owner/occupants to buy?
Perhaps Shiller is just tiring of the wait for the lag to end.
In re: Shiller, I think someone finally showed him the effect of the mortgage-interest deduction and cap-gains shield. Residential housing can be preposterously overpriced, but to a taxable investor with something approaching a decent career, the government makes it all but impossible not to buy anyway. There is no other equivalent investment at virtually any house-price level. I wrote a toy monte-carlo simulation; even at the 2005 peak, about 40% of the time you came out ahead buying *in San Francisco*, and that was assuming housing versus step-up basis for your heirs, not cashout..
That said, I think someone else may have showed the NAHB the census’s vacancy data. Like product inventories for the energy markets, they should be a primary driver of prices. Of course, barring ethanol, the government does not subsidize gas and crude. That does make a big difference.