The problem is the oversupply

July 26, 2007

Eric Landy of Morninstar tries to help us think through the cycle in the homebuilding sector – and when we might expect to see a turnaround. And it isn’t easy.

TWST: As you look at the business, where do we stand in the cycle? Is that not
important?

Mr. Landry: It’s extremely important. The problem is it’s not easily figured
out. If it were, the industry wouldn’t be saddled with all the extra inventory.
That said, it’s helpful to take a look where the industry’s been in terms of
supply and demand to get a view of what it’s up against.
It looks to us like the country as a whole is currently saddled with at least a
1 million-unit oversupply of housing units.
How do we arrive at that estimate? There are several avenues, the simplest of
which is to look at the housing vacancy survey put out quarterly by the Census
Bureau. Currently there are about 2.2 million vacant owned houses for sale, or
2.8% of total owned units. That’s an all-time high. We have never seen anything
like it before, neither on an absolute nor a percentage basis. Looking back a
couple of decades, the series has averaged a vacancy rate of about 1.5%-1.65%.
If you assume that’s equilibrium and do the simple math, you get somewhere
between a 900,000 and 1 million-unit oversupply. This doesn’t include homes for
sale that are currently occupied, so it may be conservative. Let’s assume,
though it’s accurate.
Going forward, household formation is the primary driver of housing demand over
the long term. Interest rates and, to a lesser extent, job formation are much
more powerful factors over the short to medium term, but both are more difficult
to forecast. Simply stated, it’s tough to build more houses than households that
are formed and expect to stay in equilibrium. So if you look at the demand side,
there are several demographers forecasting that somewhere between 1.3 and 1.6
million households will be formed annually over the next decade. This is an
increase from the prior 10 years due to the fact that echo boomers will be
entering prime headship age and strong immigration. Add to that an estimate for
net removals (through disasters like fire, hurricanes as well as dilapidation
and teardowns) plus demand for second homes, and you get total housing demand of
somewhere between 1.7 and 2.2 million units annually.
Starts peaked at almost 2.1 million in 2005, fell to about 1.8 million in 2006
and probably will be down another 20% to 30% this year. So if demand is at the
high end of that 1.7 to 2.2 million range, and current production is somewhere
around 1.5 million, you can see that oversupply should be soaked up in a
relatively short period. If demand is on the lower end of that scale, and
production rates stay where they are, it’s going to be a squishy market for
quite some time.