Inferred Housing Cycles
I was sifting through the data that I have on the local real estate
market here in the Inland Empire and I started to wonder when the homes
that were on the market were built. So I put together some graphs. It
turns out that the homes that are on the market were most commonly (the
mode for you statisticians) built in 2005.
This first graph
shows the number of listings on 7/30/2008 by the year built. I kept the
vertical axis set to 120 so that it was easier to compare the graphs.
Click image for larger view
My
data does not indicate anything in particular. However, it does appear
that the year that the homes that are on the market were built tend to
have peaks and valleys. For example, a sample of data taken on March
30, 2009 with a 30 day window indicates that there is a peak of homes
built in 1979, then a relative steady pace from 1985-1989 with a drop
in 1987, then a rise again until we peak at 2005.
Keep in mind
that all of the years shown are for homes that were on the market
during the sample period. So this data does not represent any
correlation with the number of homes built during a given year. It
represents a sample of listings with a change in status within 30 days
of the sample period.
This second graph show the number of listings by year built on 6/29/2009.
Click image for larger view
Why would there be such a pattern of peaks and valleys. I
sampled the data set for several different months. I went as far back
as July 30, 2008. The actual data is a little different, but the peaks
and valleys are roughly the same.
So the data begs the question.
Why are homes built in 1987 less likely to be listed than homes built
in 1988? Why are so many homes that are built in 2005 on the market?
I
suspect, but can't seem to find anything to correlate the data, that it
may be related to historical housing booms and busts. It is my theory
that during the peak of the housing boom that the quality of
construction may decline because builders are under pressure to
complete projects at an ever increasing pace to keep up with demand. As
the boom turns to bust, they are free to spend "quality time" building
a structure.
One theory that I also considered was related to
financing of the homes. Homes that were purchased in 2005 may be
financed with Adjustable Rate Mortgages that are beginning to reset.
However, this does not explain the peaks and valleys for the older
homes. All of the data represents current listings, not the number of
homes listed in that year.
Another theory may be purely related to the number homes built
during a particular year. If lots of homes were built during 1986, but
not so many in 1987, then it stands to reason that an equal percentage
of each year will result in a higher count during boom years and a
smaller number during less that active years. Lots of homes were built
in 2005. During the period of 2006-2009 less homes were built. So
perhaps the data is just a reflection of the building cycle?
Take a look at the two graphs. What is your theory on what the data means?
Note: The data for these graphs were all in the 92336 area code. If a
larger geographical area is sampled, then the data may be more uniform.
Search the Inland Empire are MLS no registration required!
Get Personalized free listing updates via emailWhat is your home is worth? No registration required.