Tuesday, April 3, 2007

Subprime Sector Notes

I finished up some research on the subprime sector -- perhaps a bit late if the market thinks this now is a non-issue (although I beg to differ & I think I have
pulled together some new insight, if one can believe it or not :)

- Ideally to calculate the possible impact one would utilize a series of detailed projections, taking into account the total amount of subprime and alt-a loans
and possible at risk loans. The best projection and discussion I've seen on possible subprime impacts is from http://www.loanperformance.com/infocenter/whitepaper/FARES_resets_whitepaper_021406.pdf (from First American Real Estate Solutions, which claims to be the number 1 source on real estate in the US, used by Moody's)

- FARES projects approximately $300Bn loans at risk over the next 5 years from interest rate resets, and, as losses will represent the loss on collateral, approximately $100Bn of total losses to the financial system, spread over 5 years.

- A weakness of the study is that only the loans made in 04 and 05 are represented, most likely significantly underrepresenting 06 and years before
03. (study completed in 2/06, but I haven't found anything of good quality newer) So one should increase the total at risk amount (unscientifically) by approximately 30-60% to account for the extra years.

- Another weakness of the study is that interest rate reset is at current interest rate levels -- all bets are off if interest rates rise significantly -- but note that much higher interest rates are not extremely likely.

- Study assumes a level of 30% from the approximate $1 Trillion in loans that will reset at risk, which the author considers "conservative" -- and walks the
reader through his methodology. I would tend to agree as the average mortage and other credit card interest payments to income is currently under 18.20% (see
http://www.federalreserve.gov/Releases/housedebt/default.htm) -- which indicates the average consumer in the US is NOT hurting by any means.

- All in all what is most interesting is that the total amount is manageable, as the total housing stock value is approximately $18-20 trillon, and value of
outstanding loans are approximately $10 trillon (according to Moody's). The defaults and losses will be spread over at least 5 years, representing in total
3% of total mortgage value (number from the study, although more like 5% taking into account other years) and losses at 1% of total mortgage loan value.

- Also demographically the subprime sector tends to be inner city -- one study of subprime loans showed that 30-40% of the borrowers were African American. In
California -- one of the largest subprime markets -- subprime loans are popular in the farming central area in addition to inner city. The defaults will hit the
inner city harder and -- if I can venture out on this -- will also contribute to the increasing disparity trend in income distribution.

- The loans will likely default and be spread over several years, 5 or more, likely leading to a stangant real estate market in the areas where subprime and
alt-a loans are popular. Perhaps a "two tiered" real estate market will develop -- inner city and other lower income areas and middle to high income class

No comments: