Monday, March 19, 2007

How bad is subprime going to get?

There is a wide range of opinions on how bad subprime is going to get. Although the ABX index seems to have stabilized, its still trading at a distressed level. Corporate bonds have widened and stock prices have fallen. Bearish economists believe that delinquent subprime loans will turn into a large increase in homes for sale, further depressing home values. This will cause consumers generally to become pinched, not just those with poor credit, and lead to a recession.

While the bearish argument is logical, I find a more moderate scenario most compelling. Consider the following.

1) According to the MBA, 13.3% of all subprime mortgages are delinquent. The all-time low is 10.3%, so at worst, there are 3% more subprime mortgages delinquent today vs. years past.

2) In 2006, 3.9% of subprime mortgages were foreclosed upon. The June 2006 delinquency figure was 11.7%, so exactly 1/3 of the delinquent loans were foreclosed upon. This ratio has been relatively consistent over the last 3 years. If that ratio holds into the future, then the number of homes for sale due to foreclosure has only increased by 1%

3) If we assume that delinquencies increase by 50% in 2007, and half of those loans are foreclosed upon (as opposed to 1/3), then the foreclosure rate would increase from 3.9% to 10.0%. That's a marginal increase of 6.1%

4) Also according to the MBA, there are about 7.5 million homes with subprime mortgages. Units, not dollars.

5) Multiply the marginal increase in foreclosures by the homes with subprime mortgages, and you get 457,500 new units are for sale.

6) According to J.P. Morgan, there are 4.1 million new and existing homes for sale. So the new foreclosures result in an 11% increase in homes for sale.

7) We know that home builders are going to scale way back in 2007. Building permits have declined by 624,000 over the last 12 months. If we assume that this represents an eventual decline in homes available for sale, there would actually be a net decrease in home supply.

8) Of course, we know that there is a serious overhang in home inventory currently. But that has been a problem for several months now. The home builders are all reducing or eliminating building homes on spec. I believe its highly unlikely that new home building outpaces new home sales over the next year.

9) So it seems as though there will be little or no net increase in homes available for sale in 2007, even under a very bearish scenario for subprime foreclosures.

10) If there is no increase in supply, how can you estimate a large and pervasive decrease in home prices? I can believe that foreclosure sales will occur at relatively weak prices, and this might depress market prices somewhat. But with foreclosures accounting for something like 10% of all sales, its hard to envision the foreclosure discount effect as creating a large and pervasive change in prices.

11) What about tightening lending standards? Won't this cause a decrease in demand from first time home buyers? Perhaps, but I think a moderate decline in interest rates will overwhelm this effect. If the Fed does cut once or twice and the 10-year stays around 4.50%, mortgage rates will fall. First time home buyers with spotty credit might be priced out of the market, but low rates will price more prime borrowers into the market.

12) Besides, nationwide the job situation remains strong. The long-term evidence is that job growth drives home sales. While we may be in a unique situation with home prices coming off a bubble, strong job growth should create a floor beyond which prices are unlikely to fall.

No meaningful increase in supply, no meaningful decrease in demand. I think the housing market will suffer at worst a mild correction.

7 comments:

Anonymous said...

Accrued Interest wrote:
12) Besides, nationwide the job situation remains strong. The long-term evidence is that job growth drives home sales. While we may be in a unique situation with home prices coming off a bubble, strong job growth should create a floor beyond which prices are unlikely to fall.

But labor markets may adjust to the subprime bust with a lag.
I would be interested in a more detailed analysis of the health of the labor market. Specifically, how much of the marginal job growth over the past few years in states like CA and AZ were real estate related ( construction, mortgage banking). I suspect the number is pretty high and that the risk in 2007 is no job growth=no home sales= recession.

Vivek Vish said...

Agreed with anonymous. Employment is a lagging indicator. I wouldn't use low unemployment as an indicator of economic health currently and in the future, although I doubt mortgage banking actually added many jobs.

Accrued Interest said...

I started writing a whole long comment to both anon and vivek but gave up and made it a new post. Thanks for the comment.

Anonymous said...

The numbers outlined in this post gave me a queazy feeling. Things are bad, and they are getting worse.

I saw an interestng post on KB homes,

http://economicdespair.blogspot.com

the basic idea is that housing market data was warning of an impending collapse since 2005.

Taking the two post together, we are looking at a very grim future for the US economy.

Accrued Interest said...

Dunno how you got that from my piece...

Anonymous said...

These comments have been invaluable to me as is this whole site. I thank you for your comment.

Accrued Interest said...

Thanks for the kind words. I'm sure we'll have more to say on subprime and CDOs.