Forum Replies Created
-
AuthorPosts
-
tickets
ParticipantPOwayseller just read the title higher in this thread
and consider the effect of the word “of” with just a slight shift in spellingParticipantGotta say I’m not happy with how the previous reply was Titled!
Participantpowayseller
I’m a jack of all trades, and master of none. Worked for a few years doing mortgage insurance modeling, a couple of years doing secondary market stuff, a couple of years doing commercial stuff. That’s why I have a pretty good feel for how subprime MBS was structured before the current boom, but don’t claim to have more than a hazy guess about how it’s structured currently. Does mean I keep up with the industry rags, National Mortgage News, Inside Mortgage Finance, etc.
The numbers I posted were for illustration only. Don’t make your financial decisions off of those – do the research. But when I was playing with Standard and Poors Levels software a few years ago I know that a portfolio of loans with even a few blemishes needed a lot of support to get AAA. A normal sort of prime pool, with FICOs in the high 600s and above and 70%-90% LTV (over 80% with insurance) started as a BBB, or maybe a single A if the scores were really good, until you added enhancements like subordination and pool insurance.
MBS structure is no great mystery. The rating agencies talk about it on their websites. Any good university library will have a copy of Fabozzi’s (one b or two in Fabozzi, I don’t remember) Handbook of Mortgage Backed Securities. It’s the reference, but like any other book in a market that moves this fast, it’s a couple of years out of date by the time the latest edition hits the shelves. But in a country where most college students have a hard time calculating percentages, a book focused on probability and integral calculus will not be a big seller.
Participantstandard and poors, moodys, and fitch all have newsletters on their websites that discuss MBS ratings. No doc definitely changes the risk status. I don’t know about recourse status. But my guess is that that shows up in servicing prices. One thing I didn’t’ go into, in part because I’m a little hazy on it myself, is recoveries from defaulted loans. Not sure what you mean about “bundling” but I don’t see any obvious reason for subsidies to flow between low and high risk borrowers.
MBS is sliced and diced into tranches in thousands of different ways. I don’t know if this is still true, but at one time the Z tranche in the subprime world was fairly small and usually held by the originator, who treated it as more or less a lottery ticket (it probably will pay next to nothing, but if times are really good ….). Don’t know if this is still the case, or if Z tranches are getting sold now. The big loser if things go south in the way of a usual cycle is the mezzanine tranche holder, and the insurer. If things go way south, then the holders of the senior tranches get hit. The $64 Billion question is “who holds all these mezzanine tranches?” No one seems to know for sure, but everyone seems to think it’s hedge funds. Probably only the hedge funds and the investment banks know for sure.
Just like no one seems to know who owns this stuff, no one seems to know what kind of return they are getting. But people have made a lot of money for decades on junk bonds. If you get a big enough interest rate you can suck up a lot of losses.
ParticipantTake 10 mortgages, each with a balance of $200,000. Put them together into a $2 million MBS. Turn the MBS into 3 tranches, a senior tranche that gets paid the first $1 million, a mezzanine tranche that gets paid the next $500,000, and a Z-tranche (often known in the business as toxic waste). So long as less than half of those 10 mortgages go bad, the senior tranche collects everything. The mezzanine tranche then starts to collect, and is OK so long as at least 8 of the 10 mortgages are good. The Z tranche is risky as hell.
To get a AAA rating on a mezzanine tranche is almost impossible, and on a Z tranche literally impossible. To get a AAA on a senior tranche the rating agencies look at how much of the MBS has been subordinated (50% is VERY good, 10% is not so good), on the loan-to-value distributions and credit scores of the underlying loans, and on the insurance. Insurance is of 2 types – the standard private mortage insurance that comes from PMI, GE, Radian, etc., and pool insurance that comes from bond market isurers. A pool of all 80% or less LTV fixed rate mortgages with FICO scores over 680 may require very little subordination to get a AAA (it will require some), a pool of 95% subprimes with option arms may require 50% subordination and pool insurance to get a AAA.
ParticipantThe article says these are AAA and AA TRANCHES. It doesn’t say they are holding the whole mortgage, just the safest tranche.
Participantcredit scoring. mortgage credit scoring models were developed in the 1980’s but only went into widespread use in the mid 1990’s. They reduced the adverse selection problem and allowed institutions to make money with looser underwriting standards. The credit expansion started the ball rolling. Low interet rates added plenty of fuel to that fire.
Participanthttp://www.fdic.gov/bank/analytical/regional/ro20062q/na/2006_summer04.html
See Chart 5. This isn’t really what you wanted – it’s just the national figure, not local figures. But it’s still interesting at the national level. Mortgage charge-offs would have to double from the end of 2005 to reach the historical average (that shouldn’t take too much longer).
ParticipantYou said your company doesn’t do neg ams, but have you seen what your competitor’s offer? On Ben’s blog a poster claimed that the 110% cap doesn’t just apply to the negative amortization, but that (at least some) lenders can demand a reappraisal and accelerate the payments if prices fall (basically, a margin call). If that’s true then a lot of borrowers will be very surprised.
Also, how well do you think pepole understand prepayment penalties? Do they realize that they should get a break on the rate if they agree to one? Do they get talked into refinancing even if they have a big prepay penalty, or a penalty that’s about to expire?
Participantinventory over sales rate IS the forward looking metric. If sales rate doesn’t change, that’s how long it will take to run off the current inventory. DOM is the backward looking metric. If 1,000,000 homes are listed, and 1 sells that month, and it took 30 days to sell that 1 house, then the backward looking metric, DOM, is 30 days, and the forward looking metric, inventory to sales, is 1,000,000 months. The huge discrepancy between the two just shows how quickly sales have fallen and inventory has risen.
-
AuthorPosts
