It just baffles me how the idea that these loans would work out spread to so many major financial institutions simultaneously. Herd behavior at its finest (worst)?
Hmmm, let's see. Prepayments are down across the board....maybe because sales are down? Maybe because a bunch of FBs owe more than they can sell the house for?
This is what happens when you speculate. All recent homedebtors, and specifically in CA, are speculators. Period.
The 40-year term thing is really mind-blowing. Per UBS, 9% of 2005 OAs had a 40-year term, but 30% of 2006 and 34% of 2007 OAs had a 40-year term.
So you've got folks who are unable to make the minimum payment based on an average payment rate of 1-2% calculated on a 40-year term. But hey! Their FICOs are higher than subprime!
fwiw, if you add in 2007 to the delinquency chart, it's even worse than 2006, just like we've seen for subprime, and the prepay speeds are trending slower for 2007 than 2006.
BoA also reports that prepayment speeds have slowed dramatically for outstanding OAs, including those with and without prepayment penalties (or with expired prepayment penalties).
But Early Payment Default is one of the biggest risks to the value of the loan, right?
So this particular stat is arguably good news for the holder(s) of the loans, at least until those slowing prepayments turn into actual deliquincies... Oh.
Banks count neg am on OA's as profit. Somehow I don't see any of this profit materializing on loans that are delinquent. If the borrowers can't even pay the principle, what hope do they have in paying the interest.
"U.S. Tosses Lifeline to Lenders Using Home Loan Banks"
"Oct. 30 (Bloomberg) -- Banks shut out of the market for short-term loans are finding salvation in a government lending program set up to revive housing during the Great Depression."
"The sales pushed outstanding debt up 21 percent to a record $1.15 trillion, an amount that may become a burden to U.S. taxpayers because almost half comes due before 2009."
What's a little more government debt for the taxpayer, a trillion here, a trillion there...
If the commentors on this thread cannot even spell 'principal', what hope do they have in understanding that historic deliquencies of O/As after the first recast are utterly useless for forecasting future performance?
Do you happen to have any statistics on the historic deliquencies of O/As after the first recast?
These BoA charts are the only things I've seen recently that compare recent to older vintages.
The thing is, there is very little "history" for this product in any meaningful sense. OA origination in the years prior to 2004 was miniscule in comparison to recent vintages, and the overall loan quality was much higher pre-2004.
That was the major mistake made on this stuff: because "historical" delinquencies were so low (on a small, carefully selected batch of loans), a bunch of fearless originators decided that they would stay that low if you opened up the product to a much wider, much less carefully selected group of borrowers.
"Old" OAs (2000-2004 originations) just didn't get old enough to get recast: they refinanced into a new one.
There's some very old data on neg am ARMs from the 80s, when they were originated by thrifts and had terrible performance problems, although I've never seen any of it online.
Banc of America is the investment/trading arm of Bank of America. IIRC from when I had the misfortune of working there (killer city, great building, awesome commute, atrocious management issues) a company that has the word "bank" in its name is or was subject to certain regulations.
I am but a lowly programmer, so I have no idea if that is still the case.
you don't get to loan money to people that can't repay it, and get it back
One thing I don't understand still is what % of this $12T in total mortgage debt is in CDOs etc vs. (for lack of a better term) "virtual money" created by fractional reserve lending.
the other thing to note is that average original credit enhancement levels for 2005 OA deals were higher versus 2006 (i'm just looking at MTAs). i would type in the actual numbers, but i'm lazy. also, virtually all OA deals were 6packs back in 04, but it shifted to most being OC in 06/07, which is not surprising given the changing quality of the loans.
"Countrywide said last week that under its new, more-conservative lending policies, 89% of the option ARM loans it made last year would no longer pass muster."
What proportion of option ARMs are Alt-A vs. subprime?
OAs are not subprime. There is no question about that.
They are best understood as a subset of Alt-A.
The "Alt-A" used by BoA for comparison purposes just means "Alt-A ARMs that are not OAs," as distinct from "Alt-A ARMs that are OAs."
The question, at least in some quarters, has been whether you could consider OAs to be prime, not whether you could consider them subprime.
The fact is, the rate of stated income on OAs is higher than for all Alt-A, and the average FICO for OAs is lower than for all Alt-A. So it makes no sense to consider them prime, as a group.
OC means deals that utilize overcollateralization as a credit enhancement (the amount of collateral exceeds the amount of liabilities issued to absorb losses). 6pack deals don't have OC, just senior/sub credit support. 6packs are normally used for better quality loans, like prime jumbos, because they don't need as much CE.
here's how moody's deals with the lack of historical data for OAs:
Moody's Mortgage Metrics'® time series simulations provide a strong tool for quantifying the impact of the
amortization gap potential across various economic environments. For each month in the life of an Option ARM loan the performance of an equivalent fully-amortizing loan with a higher LTV at origination can be modeled. In this manner, the time series performance of a collection of amortizing loans can be used to simulate the performance of an Option ARM loan over its life. As a simple illustration let's take the example of an Option ARM loan with an original balance of $70.0 (see Figure 7 below). Due to negative amortization the loan reaches a balance of $71.0 by the end of the first year. This loan's performance at month 12 can be compared to that of a fullyamortizing loan that was originated at the same time as the Option ARM loan and has similar credit features except for a higher balance of $71.5 at origination. In twelve months the second loan amortizes down to $71.0 and at this point is similar to the Option ARM loan. Similarly, in year two the Option ARM loan reaches a balance
of $73.0 and its performance at month 24 is equivalent to that of a fully-amortizing loan that had a balance of
$73.9 at origination and has amortized down to $73.0 by the end of year two. Using this approach on a periodby-period basis, the Option ARM loan's periodic and lifetime default, prepayment, and severity performance
can be projected using our deep analysis of conventional fully-amortizing loans.
The simulated loan analysis incorporates both the original LTV as well as the current LTV. Our analysis of the historical performance of loans with different amortization paths, such as 15-year fully-amortizing loans and 30-year loans, has shown that all else being equal, the LTV at origination is a stronger predictor of future performance
than current LTV. Therefore, Moody's further refines the simulated loan approach to give original LTV a
greater weight than current LTV.
Notice that CFC is going down. I just don't understand their guidance for fourth quarter and beyond. They said that by drastically tightening their standards in a sharply slowing market, they are suddenly going to make a lot of money. It's true, that if they can survive until there is a recovery, they will have fewer competitors. But they need to hold out for several more years, not just a few months.
"Banks fall in the breeze
Like woodland leaves swirling
To a quiet death"
Anyway, it is mindboggling to watch this. Gee, duh, why is it when people have a monthly mortgage payment equal to or greater than their monthly pay, there is a problem? Durr... And why can't we just assume everyone will pay off their unpayable loan and book it all as profit? Why can't I use that tactic and book my lifetime estimated salary total right now and just go buy whatever I want? Argh!!
Is there any data on how long it takes the average OA to reach its ceiling, assuming that minimum payments are made (which seems to be what the vast majority of people with OAs are making)?
Moody's Mortgage Metrics'® time series simulations:
when a biased estimator is better than nothing at all
Our analysis of the historical performance of loans with different amortization paths, such as 15-year fully-amortizing loans and 30-year loans, has shown that all else being equal, the LTV at origination is a stronger predictor of future performance than current LTV.
Per lots of analysis here at CR, original LTV would seem to be an unconditionally strong predictor, given other loan factors. But I wonder whether its relative strength is what Moody's says it is in the current regime, where you might have a stretched borrower changing LTV on a jumbo loan in a falling market.
One thing that burns me up personally is that there IS prior data on what happens when normal-to-marginal borrowers get access to negatively amortising loans during a housing bubble.
We had them here in Australia during the last housing bubble, the one that burst in the early 90's. They were called "low start loans", and a lot of them were in fact issued or insured by government institutions with the specific objective of getting people "on to the housing ladder".
These loans died a horrible death when prices dipped, complete with the usual exhortations for debt forgiveness, loan modification etc etc. Almost inevitably, they're back again now over here.
ozajh, my guess is that almost none of the worldly, sophisticated geniuses on Wall Street knew of your Australian experience. And the few that did kept quiet.
First time I heard something like this. I am shocked.
OT - Oooof - O'Neal won't receive severance, will keep invested stock and option awards.
AKA - here is your boat anchor, catch.
hey, English major! You call them "Bank", but the source says "Banc". oh, i'm so confused!
??????
So, going back to the chart of resets/recasts by month it looks like 2011 is the best hopes for recovery, barring any other fiancial/world crisi (ha!)
"prepayment speeds have slowed dramatically"
No sale, huh? I forsee a bear market in used Harleys and jetskis.
Figs 2.3 and 2.5 say "Alt-A.." on the Y axis. Is that corrent? Or are these supposed to represent all POAs?
Wow, those results are so "surprising".
It just baffles me how the idea that these loans would work out spread to so many major financial institutions simultaneously. Herd behavior at its finest (worst)?
Hmmm, let's see. Prepayments are down across the board....maybe because sales are down? Maybe because a bunch of FBs owe more than they can sell the house for?
This is what happens when you speculate. All recent homedebtors, and specifically in CA, are speculators. Period.
At what point can we stop calling this a "subprime problem"?
So much for the great W economy. No matter what the mortgage type, nobody can pay 'em. But you know, it really is different this time!
Figs 2.3 and 2.5 say "Alt-A.." on the Y axis. Is that corrent? Or are these supposed to represent all POAs?
It's supposed to be POAs. That's not the only major mislabeling in the document, but I didn't snip out the other mislabeled charts.
We'll consider my calling them a "bank" instead of a "banc" to be a childish quid pro quo.
looking at the %w/ silent 2nds and the % 40yr since 2001, the numbers for OAs and subprime arms are very similar.
bakon dreamz
fake "bakon dreamz"?
Those curves have an awesome rate of change. I wouldn't be surprised if the default rate was part of the trigger for the August "credit crisis".
No matter how you sliced it, the thing is mirage. You don't get to loan money to people that can't repay it, and get it back.
The 40-year term thing is really mind-blowing. Per UBS, 9% of 2005 OAs had a 40-year term, but 30% of 2006 and 34% of 2007 OAs had a 40-year term.
So you've got folks who are unable to make the minimum payment based on an average payment rate of 1-2% calculated on a 40-year term. But hey! Their FICOs are higher than subprime!
fwiw, if you add in 2007 to the delinquency chart, it's even worse than 2006, just like we've seen for subprime, and the prepay speeds are trending slower for 2007 than 2006.
Tanta or CR,
Do you happen to have any statistics on the historic deliquencies of O/As after the first recast?
Thanks,
Kerpowski
BoA also reports that prepayment speeds have slowed dramatically for outstanding OAs, including those with and without prepayment penalties (or with expired prepayment penalties).
But Early Payment Default is one of the biggest risks to the value of the loan, right?
So this particular stat is arguably good news for the holder(s) of the loans, at least until those slowing prepayments turn into actual deliquincies... Oh.
Wow,Affordability Products ain't what they used to be.
Banc? Were these charts produced in India?
But hey! Their FICOs are higher than subprime!
yeah! the DTI is also lower!
Banks count neg am on OA's as profit. Somehow I don't see any of this profit materializing on loans that are delinquent. If the borrowers can't even pay the principle, what hope do they have in paying the interest.
You don't get to loan money to people that can't repay it, and get it back.
Sure you can, from the gubbermint bailout.
A bit OT and apologies if this link has already been posted.
U.S. Tosses Lifeline to Lenders Using Home Loan Banks (Update1) - Bloomberg.com
"U.S. Tosses Lifeline to Lenders Using Home Loan Banks"
"Oct. 30 (Bloomberg) -- Banks shut out of the market for short-term loans are finding salvation in a government lending program set up to revive housing during the Great Depression."
"The sales pushed outstanding debt up 21 percent to a record $1.15 trillion, an amount that may become a burden to U.S. taxpayers because almost half comes due before 2009."
What's a little more government debt for the taxpayer, a trillion here, a trillion there...
How about some option arm haiku?
Bank neg am income
Not really going to be paid
Banks burned badly
If the commentors on this thread cannot even spell 'principal', what hope do they have in understanding that historic deliquencies of O/As after the first recast are utterly useless for forecasting future performance?
Bank neg am income
Not really going to be paid
Taxpayer burned badly
Do you happen to have any statistics on the historic deliquencies of O/As after the first recast?
These BoA charts are the only things I've seen recently that compare recent to older vintages.
The thing is, there is very little "history" for this product in any meaningful sense. OA origination in the years prior to 2004 was miniscule in comparison to recent vintages, and the overall loan quality was much higher pre-2004.
That was the major mistake made on this stuff: because "historical" delinquencies were so low (on a small, carefully selected batch of loans), a bunch of fearless originators decided that they would stay that low if you opened up the product to a much wider, much less carefully selected group of borrowers.
"Old" OAs (2000-2004 originations) just didn't get old enough to get recast: they refinanced into a new one.
There's some very old data on neg am ARMs from the 80s, when they were originated by thrifts and had terrible performance problems, although I've never seen any of it online.
Re: bank vs. banc
Banc of America is the investment/trading arm of Bank of America. IIRC from when I had the misfortune of working there (killer city, great building, awesome commute, atrocious management issues) a company that has the word "bank" in its name is or was subject to certain regulations.
I am but a lowly programmer, so I have no idea if that is still the case.
Cheers,
prat
you don't get to loan money to people that can't repay it, and get it back
One thing I don't understand still is what % of this $12T in total mortgage debt is in CDOs etc vs. (for lack of a better term) "virtual money" created by fractional reserve lending.
a company that has the word "bank" in its name is or was subject to certain regulations
ah, as the Japanese say, "naruhodo!".
alternatives from the meeting:
Banque of America
Banky of America
Bhank of America
Bankof America
Ban Kov America
Bänk øf Âméricä
the other thing to note is that average original credit enhancement levels for 2005 OA deals were higher versus 2006 (i'm just looking at MTAs). i would type in the actual numbers, but i'm lazy. also, virtually all OA deals were 6packs back in 04, but it shifted to most being OC in 06/07, which is not surprising given the changing quality of the loans.
"Countrywide said last week that under its new, more-conservative lending policies, 89% of the option ARM loans it made last year would no longer pass muster."
Can Countrywide Fulfill Its Profit Vow? - WSJ.com
Tanta
What proportion of option ARMs are Alt-A vs. subprime? Are any option ARMs included in the category of Alt-A ARMs?
I am trying to figure out where OAs would be included in Loan Performance pool data, which is divided into jumbo, Alt-A, and subprime.
Thanks
6packs versus OCs? Huh?
Troy, you forgot "Bankrupt America".
What proportion of option ARMs are Alt-A vs. subprime?
OAs are not subprime. There is no question about that.
They are best understood as a subset of Alt-A.
The "Alt-A" used by BoA for comparison purposes just means "Alt-A ARMs that are not OAs," as distinct from "Alt-A ARMs that are OAs."
The question, at least in some quarters, has been whether you could consider OAs to be prime, not whether you could consider them subprime.
The fact is, the rate of stated income on OAs is higher than for all Alt-A, and the average FICO for OAs is lower than for all Alt-A. So it makes no sense to consider them prime, as a group.
6packs versus OCs? Huh?
OC means deals that utilize overcollateralization as a credit enhancement (the amount of collateral exceeds the amount of liabilities issued to absorb losses). 6pack deals don't have OC, just senior/sub credit support. 6packs are normally used for better quality loans, like prime jumbos, because they don't need as much CE.
"Bank neg am income
Not really going to be paid
Taxpayer burned badly"
Broward Horne:
that's not a haiku.
(actually the first version wasn't either, since it was 5-8-5, unless you take artistic license somehow)
you can change it to this
Bank neg am income
Not going to be paid back
Taxpayer burned
here's how moody's deals with the lack of historical data for OAs:
Moody's Mortgage Metrics'® time series simulations provide a strong tool for quantifying the impact of the
amortization gap potential across various economic environments. For each month in the life of an Option ARM loan the performance of an equivalent fully-amortizing loan with a higher LTV at origination can be modeled. In this manner, the time series performance of a collection of amortizing loans can be used to simulate the performance of an Option ARM loan over its life. As a simple illustration let's take the example of an Option ARM loan with an original balance of $70.0 (see Figure 7 below). Due to negative amortization the loan reaches a balance of $71.0 by the end of the first year. This loan's performance at month 12 can be compared to that of a fullyamortizing loan that was originated at the same time as the Option ARM loan and has similar credit features except for a higher balance of $71.5 at origination. In twelve months the second loan amortizes down to $71.0 and at this point is similar to the Option ARM loan. Similarly, in year two the Option ARM loan reaches a balance
of $73.0 and its performance at month 24 is equivalent to that of a fully-amortizing loan that had a balance of
$73.9 at origination and has amortized down to $73.0 by the end of year two. Using this approach on a periodby-period basis, the Option ARM loan's periodic and lifetime default, prepayment, and severity performance
can be projected using our deep analysis of conventional fully-amortizing loans.
The simulated loan analysis incorporates both the original LTV as well as the current LTV. Our analysis of the historical performance of loans with different amortization paths, such as 15-year fully-amortizing loans and 30-year loans, has shown that all else being equal, the LTV at origination is a stronger predictor of future performance
than current LTV. Therefore, Moody's further refines the simulated loan approach to give original LTV a
greater weight than current LTV.
thx, Bakon. I am familiar with overcollaterization. It's just that OC, to me, is a crap show on FOX.
Whoa - on that Moody's item. A quote from Fargo came to mind...
"I'm not sure I agree with you a hundred percent on your {deep analysis} work, there, {Moody}".
...at month 12. check.
...at month 24. check.
...at month 61..doh!
Bank neg am income
Not going to be paid back
Taxpayer burned
This still isn't a haiku, because there is no reference to nature. Banking neg am as income is not a natural act.
This is senryu.
Notice that CFC is going down. I just don't understand their guidance for fourth quarter and beyond. They said that by drastically tightening their standards in a sharply slowing market, they are suddenly going to make a lot of money. It's true, that if they can survive until there is a recovery, they will have fewer competitors. But they need to hold out for several more years, not just a few months.
just for the record, i hate haloscan. look at that paste job. atrocious. sigh.
How's this for haiku?
"Banks fall in the breeze
Like woodland leaves swirling
To a quiet death"
Anyway, it is mindboggling to watch this. Gee, duh, why is it when people have a monthly mortgage payment equal to or greater than their monthly pay, there is a problem? Durr... And why can't we just assume everyone will pay off their unpayable loan and book it all as profit? Why can't I use that tactic and book my lifetime estimated salary total right now and just go buy whatever I want? Argh!!
The more I look at these plots, the more I realize that these curves from 2004 -2007 are going into the history books.
The caption has yet to be written.
Is there any data on how long it takes the average OA to reach its ceiling, assuming that minimum payments are made (which seems to be what the vast majority of people with OAs are making)?
How's this for haiku?
"Banks fall in the breeze
Like woodland leaves swirling
To a quiet death"
Nice... how about this:
"Solemn banker sighs
Autumn leaf to street below
It's neg am season"
Looking to the future:
Winter snows falling
Ice crystals on broken panes
Homes no longer homes
Moody's Mortgage Metrics'® time series simulations:
when a biased estimator is better than nothing at all
Our analysis of the historical performance of loans with different amortization paths, such as 15-year fully-amortizing loans and 30-year loans, has shown that all else being equal, the LTV at origination is a stronger predictor of future performance than current LTV.
Per lots of analysis here at CR, original LTV would seem to be an unconditionally strong predictor, given other loan factors. But I wonder whether its relative strength is what Moody's says it is in the current regime, where you might have a stretched borrower changing LTV on a jumbo loan in a falling market.
One thing that burns me up personally is that there IS prior data on what happens when normal-to-marginal borrowers get access to negatively amortising loans during a housing bubble.
We had them here in Australia during the last housing bubble, the one that burst in the early 90's. They were called "low start loans", and a lot of them were in fact issued or insured by government institutions with the specific objective of getting people "on to the housing ladder".
These loans died a horrible death when prices dipped, complete with the usual exhortations for debt forgiveness, loan modification etc etc. Almost inevitably, they're back again now over here.
ozajh, my guess is that almost none of the worldly, sophisticated geniuses on Wall Street knew of your Australian experience. And the few that did kept quiet.
Referenced and comments on this post were made here.
Winter (Economic and Market) Watch » More on the Emerging Pay Option and Alt A Fiasco