Is the Subprime Meltdown Contained?

There have been numerous reports and comments by people far more in the know than myself who have proclaimed the subprime meltdown as a self-contained anomaly that will not affect the much larger prime marketplace. If it doesn’t spread it isn’t a huge deal, because while it is a shame to see businesses close and borrowers lose their home, it is a necessary correction and one that the overall mortgage market can handle.

I have been a proponent of the view that the average credit rating of the borrower is somewhat of a red herring. I have argued in previous posts that while credit is part of the equation a large determinant of whether a person will default is capacity. The capacity for the person to repay will drive a large portion of loans in default or delinquent – regardless of credit score.

There are signs that banks are now considering that the meltdown will move to the prime market (having already conceded that it will spread to Alt-A).

To look at my hypothesis I took all of the loans that my company funded in 2005 and 2006 and broke them down by credit range and then took the average top and bottom ratios from the loans. These ratios represent the percentages of their housing payments and total monthly obligations respectively, compared to their gross monthly pay – an indication of the financial burden put on the borrower due to the loan.

Here they are:

2006

Credit Range	Avg. Credit	Avg. Top	Avg. Bottom	% of Production
850-680	724.35	29.78	37.2	26.03
680-620	653.19	33.42	41.56	36.53
620-560	591.58	33.15	42.03	26.84
560-500	525.25	38.67	46.44	9.36

2005

Credit Range	Avg. Credit	Avg. Top	Avg. Bottom	% of Production
850-680	734.23	32.05	39.72	43.98
680-620	647.86	31.6	38.61	26.59
620-560	597.6	35.87	44.38	19.33
560-500	510.59	33.94	41.82	7.8

After looking at this data there are a couple of interesting conclusions that I think are worth talking about.

1. In 2006, the average top and bottom DTI ratios for people with average credit scores of 653 and 592 are nearly identical. The loan burden is the same for these two groups. This should mean that they face similar risks of default due to the inability to meet their housing obligation. The only thing that would make the 592 avg. group more susceptible to default is the 653 group’s propensity to better manage their debt. However, in a scenario where ARMs are adjusting 3-4%, how large a factor does "better debt management" play in protecting the people with better FICOs?

2. In 2006, there is a significant difference between people with an average FICO of 724 and the next step down in terms of DTI. This speaks to the idea that the prime borrowers have less burden and therefore can handle an adjustment to their mortgage better. This seems to only hold for those with the best credit.

3. In 2006, the people with the worst credit have the largest DTI burdens. This correlates well with a larger default rate in this segment as ARMs adjust. They are also taking on a larger burden compared with the similar group in 2005 – nearly 5% more debt burden. This could be tied to many factors including looser underwriting guidelines and larger loan amounts.

*Please note that the percentage of production numbers represent a business strategy shift for our company from 2005 to 2006 and don’t necessarily point to a shift in lending practices or applicants year-over-year.

**Please note further that this is just one company’s data.

Next, I will look at what percentage of each group consists of ARMs vs. traditional loans vs. exotic mortgages (I/O and pay-option).