I thought it would be interesting to look at the current downturn in housing in the United States and compare it with past housing led recessions. The US has officially been in a recession since early 2008 but to those in the building industry, I am sure that it feels far more like a depression than a recession. The contraction in the building industry started in early 2006. January, 2009, marks the third year of declines. Any time that supply rises faster than demand, supply has to shrink in order to regain an equilibrium. By that measure, It is clear that the industry is paying the price for years of oversupply. The only way a lasting recovery can take place is to significantly reduce the inventory of new homes.
To see how the current downturn measures against previous housing downturns, I looked at new housing starts. Looking at the United States monthly data to 1959 allowed me to calculate the long term average in housing starts. That long term average from 1959 to 2008 is 1534 units. The data also pointed to a number of times that housing went through similar oversupply situations that then led to lower than average housing starts for years.
The most recent data for seasonally adjusted housing starts was in December 2008 and came in at 550,000 units. What I found interesting about this number is the magnitude of the dropoff in housing starts in the last number of years. From data, going back to 1959, the dropoff is completely without precedent.
Since no data going back 60 years could tell a story like the one that is currently unfolding, I decided to look at housing starts in the Great Depression and see how they compare to current starts. Since it is important to look at the oversupply number, I took the peak housing starts year. The peak housing starts number, in both the Great Depression and today, happened prior to the general economy entering a recession. Housing in both cases was a leading indicator of the recession/depression.
The Great Depression:
Housing starts peaked in 1925 at 900,000 starts. Four years later in1929, housing starts had fallen to 509,000 . By 1933 and the bottom of the cycle, housing starts measured only 93,000.*
Interestingly, it took seven years or until 1940, for housing starts to pass their 1929 levels of over 500,000 units.
Because the population of the United States is considerably larger today than in the 1930s, I thought it would be useful to compare housing starts per person. The population of the US was 123,202,624 in 1925.
Housing starts per person at the peak of the market in 1925 were 0.00731 starts per person.
Four years later, in 1929 housing starts had fallen to 0.0041 starts per person and by the bottom of the market in 1933 housing had fallen to a low of 0.00076 starts per person. Housing starts during the Great Depression fell from peak to valley from 900,000 units to 93,000 units or approximately 90%
By this measure, the recent bubble or peak of the cycle was actually worse than it was during the Great Depression. It also helps explain why this transition is truly unlike past recesions. The bubble was bigger. There was/is more inventory to work through.
The million dollar question is how long? When is the bottom?
So much of what happens next depends on the fiscal policy response. The deflation/inflation debate.
But by looking at the Great Depression, we can draw parallels to when we can expect a bottom in housing in the current cycle.
As I said previously, housing starts fell by 90% to 93,000 units at the low of the Great Depression. If we were to assume that we were to reach the lows of 1933, housing starts per person would translate into a current day housing start number of 213,000.
I believe it is highly improbable that we see housing starts fall to that level. Why?
The main difference between the Great Depression and today is in terms of housing starts is the speed of the decline. Four years after the peak of the cycle in 1925 housing starts were still at .0041 starts per person or 509,000 units. That would be the equivelent of building 1,165,054 new units today. (or using the same 4 years - a higher number today and 1,165,054 units in 2010)
This month marks three years from the peak of the housing cycle, and the speed of the reduction in building has been dramatic. The most recent data implies 0.00196 housing starts per person or 550,000 units.
The speed and the depth of this recession/depression in housing is also without precedent.
To all those that have lost their jobs through this devastating cycle, I am sorry.
To those friends still fighting: Hang in there: The current market is exactly what is needed to ensure that the current market does not last as long or go as deep as the 8 years through the late 20′s and early 30s.
The good news is that housing starts are much closer to the bottom than the top. A bottom in housing, and subsequent recovery will likely point to a recovery elsewhere in the economy. Hopefully that is sooner than later!!!