Felix Salmon

No end in sight to the housing bust

By Felix Salmon
May 26, 2009

The done thing when the Case-Shiller index comes out is to look first at the first derivative — how fast is it falling? Then people look at the second derivative — is the rate of decrease slowing down or speeding up? And if there’s no optimism there, you can always find it somewhere. The official press release leads with a graph of the first derivative over time, and then adds this:

“This is the second month since October 2007 where the 10- and 20-City Composites did not post a record annual decline. Based on the March data, however, we see no evidence that that a recovery in home prices has begun.”

I don’t think we needed an index to tell us that. But it is worth taking a step back and looking at the level of the index, rather than just its rate of change:


This is the Composite-10 index; the National and Composite-20 indices are similar, and in general show house prices at their levels from 6-7 years ago; all the same, the length and severity of the drop in house prices is still just a fraction of what we saw on the way up: we had a ten-year boom from 1997 to 2007, and there’s no particular reason why the bust shouldn’t last just as long — especially given the natural stickiness of house prices on the way down.

And what of stories announcing a “new frenzy” of house-buying in Phoenix, poster-city for the housing bubble? I think this could be a sign of a real two-way market developing, with the number of buyers approaching the number of sellers. That’s good for price transparency, but it doesn’t tell us anything about the future direction of house prices: liquid markets can fall just as easily as they can rise. And, in this case, probably will.

4 comments so far | RSS Comments RSS

Agreed… this has a long way to go down.
If the stock market has already retraced itself back to 1999 levels, housing too should be back to those levels i.e. 1999 prices just below 100 on the index.

Which means we are halfway into it…1/3rd decline done, just as much to go. How long it would take to reach there is debatable and maybe it bottoms out near 125 on the index due to the massive govt intervention.

But if i were a buyer, I would wait through this year atleast.

Posted by Andy Rebeiro | Report as abusive

I am in Phoenix and to disclose, I am not a real estate agent, so I do not care if you are buying or selling.

Right now there are 26,000 listings for homes in the entire metro area. This is down by over 20,000 in only a few months.

Right now there are 17,000 homes pending sale in the metro area. This is huge.

We have gone from 17 months supply to under 6, in a very short amount of time.

Homes are now at 1991 median prices and the most affordable in decades. they are selling for 50% of replacement value.

I can only laugh at the people saying housing will fall another 30% or so. The Fed is putting in a floor with rates and soon to be inflation, inventory is decreasing dramatically, and now 60% of all listings are ‘normal’ sales – not foreclosures. This was 10% only a few months ago.

We fell the hardest and fastest, so common sense says we’ll rebound first as well. And that’s what is going on.

Posted by joe realist | Report as abusive

“Liquid markets can fall just as easily as they can rise”

And, of course, they can rise just as easily as they can fall. Isn’t that the whole point when people cite the Phoenix situation?

As long as the market was illiquid, either sellers or buyers would have to change their prices, and since housing prices are known to be sticky on the way down, we could reasonably conclude that the sellers would change and that prices would fall. When the market is liquid, there is no longer any presumption for the direction of prices. What, then, is your basis for concluding that prices “probably will” fall even in markets that are now liquid?

It seems to me you’d make a better case by arguing that the “two-way market” is far from becoming universal. Even in Phoenix, it’s just the low end that’s liquid.

But there are several reasons “that the bust shouldn’t last just as long” as the boom. First, there’s the inflation that has taken place since the boom began. (Why should housing prices not be subject to the same inflation as everything else?) Second, there’s the net rise in average real incomes since the boom began. And third, there’s the fall in mortgage rates since the boom began.

To set off against these reasons, there’s the net tightening of mortgage credit over the same period. Basically, for someone who just wants a place to live and isn’t trying to time the bottom of the market, houses are a much better deal now than they were in 1997. It’s just that there are fewer people who are financially able to take advantage of that deal. Since we don’t know how much improvement there will be in the availability of mortgage credit in the near future, or whether it will end up still appreciably tighter than it was in 1997, or what will happen to mortgage rates in the mean time, it’s hard to tell what direction the fundamentals are pointing, but I wouldn’t assume they’re pointing downward.

Fundamentals aside, I will note that most of the market is still illiquid, which suggests that prices will go down from where they are now. I am, however, inclined to take any signs of improved liquidity as a distinctly positive sign.


From 1987 to present, the Case-Shiller Composite-10 has increased 4.06% per year on average. The CPI has increased 2.97% per year over the same period, for a whopping 1.09% difference in compounded annual growth rates.

An insight like this is completely unavailable from your chart, however, due to: (1) your use of a linear y-axis scale instead of log scale; (2) the lack of a trendline; (3) failure to include a baseline inflation comparator like CPI.

You may be absolutely correct in saying that the housing crash has further room to run, but you’re not doing anyone a service by using crappy charts to support that thesis.

Posted by Matt | Report as abusive

Post Your Comment

We welcome comments that advance the story through relevant opinion, anecdotes, links and data. If you see a comment that you believe is irrelevant or inappropriate, you can flag it to our editors by using the report abuse links. Views expressed in the comments do not represent those of Reuters. For more information on our comment policy, see http://blogs.reuters.com/fulldisclosure/2010/09/27/toward-a-more-thoughtful-conversation-on-stories/