by Calculated Risk on 5/07/2008 12:09:00 PM

Brett Arends at the WSJ asks: Is Housing Slump at a Bottom?

[The following chart] from Wellesley College Prof. Karl E. Case, one of the leading experts on the housing market in the country. And it suggests we may be at, or near, the bottom of the housing crash.

... new housing starts have at last slumped below the seemingly magical one million mark. That happened in March. Every time that has happened in the last 50 years, it proved to be the bottom of a recession.

"It is really remarkable how much where we are today looks like the bottom we've had in the last three cycles," Mr. Case says. "Every time we've gone below a million starts, the market has cleared at that moment."

First, I think any article discussing the housing "bottom" should start by defining what they mean by "bottom". Are they talking about starts? New home sales? Residential investment? Housing prices? Or some other metric?

Most people think of the bottom in terms of price, and in most housing busts, starts, residential investment, and new home sales all bottom long before housing prices bottom. The linked article seems to confuse a bottom for housing starts with a bottom for housing prices, and that is incorrect.

Second, we can write the supply side of the equation as:

Supply = new units added - units sold + existing units for sale. Looking at just housing starts provides only one portion of the equation (this leaves out rental units too - a substitute product).

Here is a graph of inventory (new and existing) for sale at year end (March for 2008):

Click on graph for larger image.

In the supply-demand equation for housing, prices will be under pressure until the total supply declines significantly. So even if housing starts are near a bottom, there will no quick recovery for starts, and prices will continue to decline, until the total inventory is reduced.