[ADDED: See comment added in December 2011 at the bottom. The original post is from April 2008.]

I have a theory that areas that are the most over-priced will “correct” more quickly than other areas.

Below is a post I had published about 3 weeks ago in an U.S. housing stocks blog called “Seeking Alpha.”


The Reverse Ripple Theory of Metropolitan Home Price Corrections

You drop a rock in water. The ripple is largest at the center and becomes weaker as it spreads out.

In the Reverse Ripple Theory of Metropolitan Home Price Corrections (you got a better name?), home prices in the outlying areas with a lot of new construction will tend to fall most and first (the splash), as it moves toward the center of the metro area, the ripple weakens and the fall in home prices is progressively less and later.

The “less” part is probably conventional wisdom among real estate geeks. The “later” part is probably not.

Will the Periphery Bottom Out First?

Buried in Appendix C of Global Insight/NationalCity‘s September 2007 issue of “Home Prices in America,” I noticed an intriguing little table barely mentioned in the text, called “Past Price Corrections.” The table is now included in recent issues, as well.

The appendix showed 79 metro area home price corrections completed since 1985. Its great stuff for real estate geeks.

Don’t miss the last two notes in Appendix C.

  • The more severe the overvaluation, the greater the subsequent declines tended to be: correlation = +0.25
  • The more severe the overvaluation, the shorter the duration tended to be: correlation = -0.31

That last note surprised me. The more a market is overvalued, the faster home prices tended to correct! That seemed counter-intuitive to me. Why would a more overvalued market correct sooner than a less overvalued market?

Anyway, let’s run with it. What if that effect occurred within a single overvalued metropolitan area as well within the history of U.S. home price corrections?

What if home prices in the most overvalued areas of metro Phoenix Arizona, my home, fell most and bottomed out first?

Most Overvalued Areas

So, where were the most overvalued areas in metro Phoenix during the recent boom?

The conventional wisdom is that areas on the periphery with lots of new homes were more overvalued, at the peak, than other areas.

The periphery became more overvalued, in my opinion, because there was a bit of a pricing oligarchy in the new home areas. The homebuilders were the oligarchy.

As the market tightened during the boom, homebuilders had the market power to increase prices in their areas more quickly than was possible in older areas where each home seller made an independent pricing decision.

When a homebuilder increased its list prices by say, $10,000, he increased the value of all the resale homes in the subdivision at the same time.

If, however, Joe Homeseller in the same subdivision increased his list price $10,000 – and the builder did not – it would have essentially no affect on the market.

In my view, prices in new home areas increased more than in other metro Phoenix areas with similarly tight markets because of the market power of the homebuilders to raise prices.

Since the peak, we have certainly seen median home prices on the periphery fall faster than in the interior of metro Phoenix.

The future will show us if these peripheral sub-markets will also tend to be the first areas to see median home prices bottom out and whether the interior sub-markets of metro Phoenix will tend to have less severe but slower home price corrections.

Hey, it’s just a theory.

The charts below compare 85086 in far north Phoenix, which includes Anthem, a newish master planned community of Del Webb (Pulte Homes), and 85051 in Phoenix, an area of homes built in the 1960s and 1970s. Charts show monthly medians for resale (not new), single family detached homes.

anthem home price trends

Obviously, the periphery (85086) started to fall first. The question is will it fall further and bottom out first?


Here is the original article.

Later the theory was referred to in an article in the San Jose Mercury News.

[End of original post.]


The theory didn’t work out like I expected as you can see in the updated chart below.

To answer my question in the final sentence of the original post, the periphery (85086) did NOT fall further and it did NOT bottom out first.

The most surprising thing to me in the updated graph is that home prices in the older, established zip code 85051 fell further.

Here’s what happened, I think.

Sub-prime loans pumped more money into 85051 in 2006 which helped home prices increase in 85051 in 2006 (while they were falling in 85086, see graph above) but it also laid the groundwork for more foreclosures and greater home price declines later on in 85051.

Home prices were significantly lower in 85051. I assume incomes were lower as well in 85051. I therefore assume sub-prime loans were much more common in 85051 than in 85086.

You have to ask, when mortgage companies started to heavily promote sub-prime loans in the second half of 2005 and 2006, who used those sub-prime loans?

A lot of people who used sub-prime loans were people who couldn’t get a loan earlier during the real estate boom because of their low incomes and poor credit scores. Now that these lower income people could get (sub-prime) loans, they wouldn’t likely go out and buy $400,000 homes in 85086, they would be a lot more likely to go out and buy $200,000 homes in 85051. Thus, 85051 ended up, I’m assuming, with a lot more sub-prime loans than 85086.

If we assume more sub-prime loans were written in 85051 that means those home buers put less money down (or no money down) and in general they had less equity than homeowners in 85086. We would expect more foreclosures later on in 85051 when prices started to fall.

So that’s my theory of why prices fell further in 85051 than in 85086, sub-prime loans.