The housing market news was been dismal last week. It has occupied my attention with two articles (here and here) plus an Instablog. Last week, David Rosenberg, Chief Economist for Canada's Gluskin Sheff, posted an interesting discussion, with the following two graphs:
In Chart 1 we see just how different the housing bubble was from any previous experience for over 40 years. When you realize that the population has almost doubled in the since 1960, normalization of houses sold per capita would produce a declining base line instead of the level one provided in the graph. See Appendix.
In Chart 2 we see the tremendous median time on the market rising to almost twice the level of the highest point in the two prior recessions. This may well go higher because single family housing starts are increasing at a time when new home sales are falling. All this will be aggravated by 2-3 million new foreclosures becoming available in 2010.
Permits, Starts and Sales
In the first graph below we concentrate on how residential building permits, single family housing construction starts and new home sales have varied since January 2006.
The long term relationships between these three metrics have permits greater than single family housing starts because the permits number includes both single family and multi-family construction. Also, the long term average for single family housing starts is larger than new home sales because starts include houses built to be rented and owner constructed housing, neither of which are intended to be sold immediately. An aside, a contractor built house on the home owner's land is included in the owner constructed category.
While the long term averages have the relationships described above, in short time intervals the relative magnitudes can get out of whack. For example, in 2Q/2008 there was a sudden spike in housing starts, which didn't get averaged out until the end of that year. Another example occurred through the second and third quarters of 2009 as new home sales exceed construction starts as builders attempted (successfully) to reduce inventory.
The graph below gives a close up for 2009.
The inventory of new homes has been reduced in 2009, but remains far above healthy levels. This is reflected in the continued increase of median time on the market for new homes in 2009, shown in the second graph of this article. The slight inventory improvement of 2009 may be wiped out very quickly because the sales and starts data are moving in opposing directions in a hurry, as seen in the graph.
As long as the recent two month moves are continued (starts rising and sales falling), there is no reason to expect home builders to be able to improve their situations.
Appendix
When Chart 1 is normalized to population (i.e., plotted per capita) the following graph is obtained
The "support" line now has a definite downward slope. This indicates that the sales level in the current recession is below what has been experienced in previous recessions. The only other time when unemployment levels were similar to now was in 1981-82, the previous low point for sales.
The reasons I can offer for why the sales level is so low this time are:
- New home prices are still above the historical ratio to median income.
- The inventory of existing homes homes is larger than in other recessions.
- The bubble took sales to levels way above other peaks, leading to over building.
- The inventory of foreclosure homes on the market at distressed sale prices is much above previous times.
- The overhang of several million more potential foreclosures that may yet come to market is strong inhibitor to sales of the more expensive new construction.
One might expect a rapid rise in new home sales coming off such a low level as has occurred in the past. Because of the unique conditions listed above, that is not likely this time.
Disclosure: No positions.
