By New Deal Democrat
Based on the increase in interest rates last summer, by now I thought that housing in general would be down about -100,000 units a year. Although the housing market in general has slowed down, that hasn't happened, and the more volatile housing starts number showed a YoY increase of more than 20% YoY in April.
When a forecast doesn't pan out as I expect, I try not to argue with the data. Instead I go back to see if I can find out what I may have missed, or what complications or nuances need to be addressed. In the case of housing, I think there is a complicating factor in the current forecast. To demonstrate that factor, I'm going to take a more detailed look at the housing market in segments. In this first installment I'll look at single-family homes. In the second, I'll look at apartment. Finally, I'll take a look at the impact of demographics.
As I'll show in a later installment, single-family houses are particularly sensitive to interest rates. While there has been a spring rebound in housing permits and starts as a whole, that rebound has been entirely confined to apartments. In fact, the recovery in building single-family homes ended at the beginning of January 2013.
First of all, here is a graph comparing the permits for single-family homes (blue) and multi-family (5 unit or more) structures for the last 5 years:
Virtually ALL of the growth in housing permits in the last year has come from multi-family units such as apartments. Contrariwise, here are single-family home permits (blue), starts (red), and sales (green). All of these have been normed to 100 as of January 2013:
There has been virtually NO growth - zip, nothing, nada - in any of these measures of the market for the building of single-family homes for the last 16 months.
Previously I have described how the housing market is very interest rate sensitive. Here's a variation on a graph I have previously run, comparing YoY housing permits for single-family homes (blue) with YoY change in 10-year treasury interest rates (red, inverted):
Note since the mid-1980s, there has been a very consistent relationship of interest rates leading single-family homes.
Now here is the updated version of that same graph focused on the last 5 years:
As in nearly all cases in the last 30 years, a YoY increase in interest rates has led to a YoY decline in permits for the building of single-family homes, although interestingly, not so deep a decline as in other cases, at least so far.
Now, let's examine permits for single-family homes (blue) vs. the Case Shiller house price index for the last 5 years:
Note that the index has continued to rise, despite the flatlining of building of single-family homes.
When we back up to a 10-year perspective, we can see that house prices kept rising after the bubble in single-family homes started to burst at the end of 2005:
The continued rise in prices almost certainly contributed to the decline in sales - and appears to be doing so again -- as to single-family homes.
Finally, let's look at the YoY% change in single-family home permits for the last 50 years:
Typically there has had to be a decline of about 200,000 in permits YoY to bring on a recession. Needless to say, we don't have that. The bottom line is that, the housing market for single-family homes has behaved close to the way I expected housing to behave based on the increase in interest rates and in house prices. Its recovery has completely halted, and permits are down about 7% from last autumn.