by Calculated Risk on 6/07/2011 08:38:00 AM

From Robbie Whelan at the WSJ: Second-Mortgage Misery

[A] report to be released Tuesday by real-estate data firm CoreLogic Inc. ... says 38% of borrowers who took cash out of their residences using home-equity loans are underwater ... By contrast, 18% of borrowers who don't have these loans were underwater.
...
Overall ... 10.9 million Americans who borrowed to buy their homes, or 22.7% of all homeowners with a mortgage nationwide, were underwater in the first quarter ...

This is a slight decrease from the 11.1 million, or 23.1 percent of homeowners with a mortgage who were underwater at the end of Q4 2010. The WSJ article notes the decline was probably due to completed foreclosures.

I'll have more when the CoreLogic report is available online.

by Calculated Risk on 6/06/2011 10:05:00 PM

As a followup to my earlier post, Housing Starts and the Unemployment Rate, Professor Amir Sufi sent me this graph of the unemployment rate and residential construction by state.

Credit: Professor Amir Sufi. "This maps the unemployment rate in a state in 2010q4 against the percentage drop in construction of new residential units from 2005 to 2010 (i.e., [new units constructed 2010 – new units constructed 2005]/new units constructed 2005)."

Click on graph for larger image in graph gallery.

Of course cause and effect can run both ways: A higher unemployment rate probably means less residential construction, and less construction would mean a higher unemployment rate (so areas with high levels of construction during the boom - like Florida, Nevada, Arizona and California - would see a higher unemployment rate during the bust).

Note: I've excerpted before from papers by Sufi. Here is a recent Fed Letter with Atif Mian: Consumers and the Economy, Part II: Household Debt and the Weak U.S. Recovery

by Calculated Risk on 6/06/2011 06:58:00 PM

Kash Mansori at the Street Light looks at some interesting data released by the BIS today on who owns the debt, and who sold the default insurance, for Portugal, Ireland and Greece: Betting On the PIGs. It ends up most of the debt is owned by European creditors, but US institutions are more exposed to a default. Kash writes:

If Greece were to default, for example, approximately 94% of the direct losses would fall on European creditors, and only 5% would fall on US creditors. However, US banks and insurance companies would have to make about 56% of the default insurance payouts triggered by such an event, while European agents would make only 43% of those payouts.
...
US and European financial institutions are likely to have very different incentives as negotiations regarding debt restructuring and reprofiling proceed. US banks and insurance companies are surely delighted with the "soft restructuring" that is currently being discussed.
...
In essence, European firms have been betting that a PIG default will happen sooner rather than later, while US firms have been betting that default would happen later or not at all.

It is interesting that European institutions were more willing to sell default insurance for debt issued by Ireland and Portugal than for Greece ... probably from experience.

by Calculated Risk on 6/06/2011 02:45:00 PM

The following is from economist Tom Lawler:

Over the weekend I “downloaded” historical data on active listings of SF homes and condos for 54 metro areas back to April 2006 from housingtracker.net (it was a pain). The historical data are monthly averages of weekly listings, and in May the 54 metro areas combined had 1,111,996 listings.

CR Note: the following graph combines two graphs from Lawler.

Below is a chart for all 54 metro areas [and] of the NAR’s estimate of the inventory of existing homes for sale over the same period (May 2011 data are not yet available).

Click on graph for larger image in graph gallery.

May listings for the 54 metro areas were down 6.8% from last May; down 10.2% from May 2009; and down 26.1% from May 2008!

Obviously, the trend in listings for the 54 metro areas is materially different from the NAR’s inventory estimates over this period. Here is a table showing April 2011 listings compared to listings in April for the previous 5 years.

% Change in Residential Listings/Existing Home Inventory, April 2011 from:
	NAR	HT 54 metro areas
Apr-10	-3.9%	-8.3%
Apr-09	-1.7%	-12.6%
Apr-08	-14.9%	-26.0%
Apr-07	-8.3%	-20.9%
Apr-06	13.3%	-1.7%

There are numerous possible reasons for these differences. First, of course, the 54 metro areas in housingtracker.net’s report are probably not as a group representative of the US market as a whole. There is a heavy representation of “bubble” markets in housingtracker.net’s list where listings soared in 2006/07 but have since fallen sharply (not true, btw, for all troubled markets, with Vegas and Reno being obvious exceptions). Second, HT’s listings may include some new homes listed for sale, which I think NAR tries to exclude.

Another big reason, however, is related to the NAR’s methodology for estimated existing home inventories. The NAR collects sales and “months’ supply” data from around 160 boards/MLS across the country, almost all of which are located in or adjacent to metropolitan statistical areas. The NAR assumes that the % changes in sales from these Boards/MLS reflect total sales across the country, and they apply these changes to “benchmark” estimates of total existing home sales derived from an analysis of decennial Census/AHS data (the last benchmarking was based on 2000/01 data, but that same methodology can’t be done with the Census 2010 data). For inventories the NAR uses the “months’ supply” data from these Boards/MLS to derive a months’ supply estimate for the nation as a whole, and “solves” for the inventory level. (Though multiplication!)

Most analysts believe – as does the NAR, by the way – that the NAR’s methodology has overstated existing home sales for the past four years or so, with the overstatement emerging no later than 2007 and growing over subsequent years. Given the NAR’s methodology of using sales and months’ supply to generate inventories, an overstatement in sales would also produce an overstatement in inventories, and a rise in the degree to which sales are overstated would result in an increasing overstatement of inventories as well. (In discussions of the NAR’s “overstatement” of existing home sales, some analysts incorrectly concluded that one implication was that the months’ supply numbers were understated. That is not correct.) Thus, inventories over the last 4-5 years have probably declined by more than that suggested in the NAR report.

There may be other differences as well. E.g., Boards/MLS do not all report active listings on a consistent basis.

But flipping back to the housingtracker.net numbers, it seems pretty clear that inventories in a large number of metro areas have declined significantly over the last several years, and have declined significantly more than a look at the NAR data might suggest.

To be sure, declining listings across most of the country is not necessarily “bullish” for housing, as home sales have also declined – that is, there has been declining demand as well as declining supply. Still, the declining supply is welcome news.

A table showing May 2011 listings compared to listings in the May of the previous 3 years is shown [below].

Active Listings, May 2011 Compared to May of Previous 3 Years, Various Metro Areas
	2010	2009	2008		2010	2009	2008
Albuquerque	-4.9%	-9.5%	-20.8%	Minneapolis	-10.9%	-11.5%	-28.8%
Atlanta	-9.9%	-21.3%	-41.6%	Nashville	-9.0%	-7.3%	-10.0%
Austin	-16.8%	-4.2%	-11.2%	New Orleans	-10.1%	-6.1%	-14.4%
Baltimore	-2.7%	-4.9%	-14.0%	New York	6.6%	6.6%	1.7%
Boise City	-30.6%	-39.9%	-47.7%	Newark	7.9%	5.2%	-3.4%
Boston	1.7%	4.4%	-20.4%	Oklahoma City	5.8%	4.6%	0.4%
Cape Coral	-33.1%	-40.2%	-46.6%	Omaha	-2.3%	9.5%	-9.3%
Chicago	-11.5%	-13.9%	-27.9%	Orange County	0.0%	10.2%	-8.7%
Cincinnati	-5.6%	11.7%	2.5%	Orlando	-23.1%	-48.3%	-57.7%
Cleveland	-3.1%	1.6%	-21.8%	Philadelphia	4.5%	6.5%	5.2%
Columbus	10.6%	14.3%	-10.8%	Phoenix	-21.0%	-31.7%	-45.8%
Dallas	-3.2%	-2.6%	-18.1%	Portland	-14.4%	-17.7%	-32.2%
Denver	-8.2%	-8.3%	-25.7%	Raleigh	-5.5%	0.1%	-3.2%
Detroit	-18.9%	-38.4%	-56.1%	Reno	8.7%	3.5%	2.2%
Edison	9.4%	10.2%	1.1%	Riverside	2.6%	-4.1%	-39.0%
Honolulu	-5.9%	-20.5%	-27.9%	Sacramento	-2.6%	8.4%	-22.5%
Houston	-5.7%	6.2%	-16.0%	Salt Lake City	-15.0%	-20.3%	-20.3%
Indianapolis	-8.4%	-4.1%	-19.4%	San Antonio	-0.8%	-1.4%	-7.6%
Jacksonville	-17.4%	-18.0%	-32.0%	San Diego	2.9%	30.8%	-15.4%
Kansas City	-3.0%	-2.4%	-22.4%	San Francisco	-3.7%	-3.3%	-29.1%
Las Vegas	4.6%	-2.8%	-2.4%	San Jose	-6.7%	-9.8%	-28.2%
Long Island	1.2%	-1.7%	-4.0%	Seattle	-15.0%	12.1%	-5.1%
Los Angeles	2.7%	6.1%	-23.1%	St. Louis	-7.4%	2.9%	-13.4%
Louisville	-4.0%	-4.1%	-11.0%	Tampa	-9.5%	-25.5%	-37.9%
Memphis	-6.7%	-12.9%	-32.7%	Tucson	-7.4%	-3.3%	-20.1%
Miami	-25.3%	-45.9%	-56.8%	Va. Beach	-6.0%	4.3%	-1.1%
Milwaukee	-8.2%	-3.1%	-10.1%	DC	-4.4%	-11.5%	-30.7%
Source: Housingtracker.net

by Calculated Risk on 6/06/2011 11:34:00 AM

An update by request: The following graph shows single family housing starts (through April) and the unemployment rate (inverted) through May. Note: there are many other factors impacting unemployment, but housing is a key sector.

You can see both the correlation and the lag. The lag is usually about 12 to 18 months, with peak correlation at a lag of 16 months for single unit starts. The 2001 recession was a business investment led recession, and the pattern didn't hold.

Click on graph for larger image in graph gallery.

Housing starts (blue) increased a little in 2009 with the homebuyer tax credit - and have declined recently, but mostly starts have moved sideways for the last two and a half years. This is one of the reasons the unemployment rate has stayed elevated compared to previous recoveries.

This is what I expected when I first posted the above graph almost two years ago. I wrote:

[T]here is still far too much existing home inventory, a sharp bounce back in housing starts is unlikely, so I think ... a rapid decline in unemployment is also unlikely.

Usually near the end of a recession, residential investment (RI) picks up as the Fed lowers interest rates. This leads to job creation and also household formation - and that leads to even more demand for housing units - and more jobs, and more households - a virtuous cycle that usually helps the economy recover.

However this time, with the huge overhang of existing housing units, this key sector hasn't been participating.

The good news is residential investment should increase modestly in 2011, mostly from multi-family and home improvement, and that will help push down the unemployment rate. But I still think the labor market recovery will be sluggish until the excess housing supply is absorbed.