Case-Shiller numbers confirm housing double dip

The Case-Shiller numbers this morning were lower than expected and confirm a housing double dip which began in July when both the Composite-10 and Composite-20 numbers peaked.  The Composite-10 is down 0.5% y-o-y and the Composite-20 is down 1.6% y-o-y.

While housing is expected to be a drag on the economy over the near term, with prices and transactions declining, and foreclosures expected to hit another record, it is unclear how much seasonality is driving the change as all numbers presented here are non-seasonally-adjusted.

As a reality check, I looked back at last Winter’s Case-Shiller reports. Here is what I wrote.

January 2010:

So, how are things looking today? Mixed.

We can speak of year-on-year gains in house prices in some markets (Dallas, Denver, San Diego, San Francisco) but on the whole, things are going in reverse. As I mentioned in the last two months, the number of markets where prices are rising has dwindled. This continues to be true.

• When the Case-Shiller index began increasing in July, 14 of 20 markets were showing an increase.
• Case-Shiller reported in August that 18 of 20 cities showed price increases.
• When Case-Shiller reported in September, 18 of 20 cities showed price increases.
• Then, in October the number turned down slightly to 17 of 20 markets.
• In November, the number really turned down. Only 10 of 20 markets rose in the data (for sales through September)
• Last month, it was 8 markets in twenty (for sales through October).
• This month it is five of twenty for data through November.

Can I give these numbers a positive spin? Sure: prices are moving more into line with reality.  They need to come down further in bubble markets like Washington D.C. where I live. Maybe then, we will see a market-clearing price.

February 2010

The December 2009 data for the widely followed S&P/Case-Shiller Indices were released this morning.  The data are showing a mixed picture. On the one hand, the unadjusted numbers are down on both a month-to-month and year-over-year basis (Composite-10 and Composite-20 respectively down 2.5% and 3.2% versus December 2008). Only four of twenty markets saw price increases. However, the seasonally-adjusted data do show a slight improvement in markets on a month-to-month basis, despite year-over-year price declines.

Below are the non-seasonally adjusted data:

What is clear from the numbers is that the markets in which prices are now doing the best are mostly the same ones that had both experienced the greatest carnage and had also experienced a prior price bubble. This includes Phoenix, LA, San Diego, San Francisco, and Las Vegas. You see yearly home price increases in California for example. Moreover, the only four markets where prices increased month-to-month were in the previously devastated bubble markets of Phoenix, Las Vegas, San Diego and Las Vegas.

The breadth of price increases has now narrowed to 4 of twenty markets. And that has to be worrying whether you are looking at unadjusted winter month data or seasonally-adjusted data. Since June, the number of markets in the Composite-20 where prices have risen has gone from 18 in June to 18 in July, 17 in August, 10 in September, 8 in October, 5 in November and 4 in December.

March 2010

The Case-Shiller Home Price Indices released this morning were a mixed bag. The data for January 2010 showed a decline of 0.7% from a year earlier. On a month-to-month basis prices fell 0.4%.

Yet nine of the 20 markets in the Composite-20 now have year-on-year price increases.

Nevertheless, only two of the twenty markets in the Composite-20 showed price gains last month. This narrowing of price inflation began in the summer of 2009 and has continued every month since then. While much of this may reflect seasonality, it is not reflective of a housing market that is about to take off.

Bottom line: US house prices are no longer in freefall. But, the market remains fairly weak. Some of the weakness right now is due to seasonal factors. However, the data for the Spring and Summer will help us gauge how much is seasonality and how much is softness in the market.

(More charts below)

I would highlight the following:

• NEGATIVE: The average national housing price indices are still above the trendline. And since prices usually revert to mean and overshoot, this suggests more decline is to come. We could see as much as 15-20% more.
• MIXED: The breadth of declines now is similar to the dynamic from last year. So clearly, much of the decline is seasonal. In a weak market where price gains are minimal, it would not be unusual to see prices decline in the Winter but advance in the summer. I should note, however, that the dip this year has come earlier and is more pronounced in breadth.
• POSITIVE: The active inventory level at 8.1 months is coming down

Bottom line: the U.S. housing market is double dipping. How much of this second dip is seasonal remains to be seen. Nevertheless, the numbers did break down starting in July – and this suggests the breakdown is not just a seasonal pattern but a fundamental weakening in demand due to less housing-oriented stimulus. As housing usually leads out of a recession, this will be a drag on growth compared to other early recovery periods.