Case-Shiller Housing Index: Regional Analysis

.
How do the different cities/states compare in the Case Shiller indexes?

I have shown a plot like this before but this shows the 20 cities tracked by the Case Shiller index (click on this or any graphics for larger views):



A few questions come to mind such as: Which cities lead and which lag the others or the 10-City composite? Which went up higher and fell further?

Different City/State Time Lags vs. the Case Shiller 10-City Composite

I looked primarily at the timing of the peak of the bubble for each city compared to the 10-City Composite. A few key observations:

• The 10-City composite peaked in June 2006.
• Boston, MA housing prices peaked 251 days earlier in Sept. '05.
• San Diego and then Detroit were next in Nov. '05 and Dec. '05 respectively.
• On the flipside, Dallas, Portland, Seattle, and Atlanta all peaked about a year later than the 10-City composite around June/July 2007. Charlotte, NC was last in August 2007

It was kinda hard to visualize so I threw this into the many-eyes website to produce a map. Here is a picture of this one:

The blue colored states are showing negative lags vs. the 10-City composite (peaked before) and the brown colored are showing positive lags (peaked after). So you can see in the blues, California, Michigan and Massachusetts housing bubbles peaking early and the Pacific Northwest, Texas, Georgia, and North Carolina peaking late. The net difference between say California and Georgia is more than two years. For California and Florida, I averaged the multiple cities' Case Shiller indexes for those states which really oversimplifies things since some cities peaked much sooner or later than others.

Which States are Dropping Most Now?

Below you can play with all the different regional maps I made with the latest Case-Shiller data but here is one more screenshot to look at which states are dropping the most. Here is the March vs. March a year ago % change:

So the darkest blue states are the ones that have dropped the most in the past year. Arizona(Phoenix) is the worst at a 36% drop with Nevada (Las Vegas) next at a 31% drop. Colorado(Denver), and Texas(Dallas) are in the best shape with annual drops around 5%.

Here is the embedded map script so you can play around with it, I have other data coded in like % gain to the peak, drop since the peak, and last month's change. Click on the chart below to activate it so you can interact and see different views it or go to the many-eyes website to play with it.

All this makes me think a bit more about forecating the housing market. What this says in general is that there were a couple of predictive housing markets (Boston, San Diego, Detroit) which peaked a year before the broader 10-city index did. So I would expect some markets to bottom also before the broad housing market does. So far, unfortunately, I see no signs of any individual Case-Shiller city index bottoming but when one does I will be sure to let you know. It might be one of the first early signs of the bottom in the overall market coming. I also may play around with another way of building a Case Shiller forecasting model by forecasting each of the cities and averaging it up to the Case-Shiller composite.

Let me know what your thoughts are and what you think!

Link to my other Case Shiller Housing Posts.
...