Does anyone know the dollar value or number of resets that are indexed to LIBOR or something other then an index directly affected by the fed funds rate?
Does anyone know the dollar value/number of resets directly affected by the fed funds rate?
My point is ... how will a fed funds rate help lower the reset rate if the reset in not indexed to the a rate that is affected by the fed funds rate?
Now, if the fed funds rate in decrease then the discount rate should follow if it continues to float against the fed funds rate. Is this the real reason for a fed funds rate decrease?
Other then stimulating wall street who will a fed funds rate cut help?
How will a rate cut affect the asset bubble in the housing market?
Kudos to you. One of the best pieces on explaining this topic I've seen. An excellent post. Has anyone told you lately this is one of the best blogs out there. Fantastic stuff.
For those who think a Fed rate cut will help I believe their general theory is the rate cut will naturally draw down all other rate indices. I'm not sold on that theory but it's out there nevertheless.
Thank you! I had wondered why the BofA chart was so much different than the CS chart from before, and had been meaning to ask this very question. I had assumed that it was just the use of proprietary info, but this makes much more sense. So it all depends on what question you ask and what you factor in, which makes using those numbers (for us outsiders, at least) a very inexact tool.
The cost of funds in the hard-hit interbank money markets hit peaks not seen for nearly a decade on Tuesday as the scramble for cash by financial groups showed little sign of easing.
An important borrowing benchmark for investors, the three-month London interbank offered rate (Libor), climbed to 6.90375 per cent in the sterling market, up from 6.89625 on Monday, leaving it 115 basis points above the Bank of Englands base rate of 5.75 per cent and at the highest level seen since November 1998.
Anyone have a link to the Sinai article about many defaults happening even before the first reset?
Does anyone know the dollar value or number of resets that are indexed to LIBOR or something other then an index directly affected by the fed funds rate?
Does anyone know the dollar value/number of resets directly affected by the fed funds rate?
I don't have the exact LIBOR percent at my fingertips (and I won't until Blogger stops hating me), but suffice it to say that almost all subprime ARMs are LIBOR-indexed, most Alt-A ARMs are LIBOR, and some prime ARMs are. If a loan isn't LIBOR-indexed, it is most likely to be treasury-indexed. (There is a small pile of weird indexes based on bank cost of funds, or COFI, but ignore that, it's a very small pile these days.)
A whole lot of Option ARMs are treasury-indexed (MTA, or monthly/moving treasury average).
You cannot get directly from Fed funds to either the cost of London interbank borrowing (LIBOR) or U.S. Treasury securities of a constant maturity of one year, six months, or one month, as the case may be for different loan types.
You also must bear in mind that ARMs reset on the schedule in the note, which will depend on the loan's origination date. ARMs that reset October 1 already have their index values determined (from 8/15 or 9/1, depending).
I did a post on that subject in the end of August, and if I could get the forking archive to load properly, I'd link to it.
Short version: Fed fund cuts might affect the price of new loans, since it affects the bank's cost of funds and thus its net interest margin. So a rate cut could create refi opportunities. It won't result in rates on outstanding ARMs dropping any time soon.
Is it feasible to create a graph showing the average (or median) monthly payment for ARMs (all or just securitized). That way.. One could just make assumptions on a range for interest rates and then graph the potential increase in the avg/median monthly payment. You could also just graph the total amount of money payed towards these mortgages over the next 12 months.
Or is the data just too variable to be mapped in this way?
It just seems like this would give a more meaningful illustration of what could really happen dollar wise.
Tanta, thank you that was very helpful. Re your comment that a rate cut might create refi opportunities, presumably these will only be available to a smaller set of borrowers than a year ago. Like, borrowers with actual income proof or say, a downpayment?
Eli, the paper on my desk tells me that subprime borrowers facing first reset in the next year ($27B/mo on avg) will get an average payment shock of 35%-45%, based on rate levels on aug. 17. The avg original loan size is like $225K.
Nice piece! I think another whammy has to do with the resets of IO's. Not only do the rates adjust at the end of the fixed rate period, but the payment is recast from an IO to one of a fully amortized loan. This can kick up the monthly payments by 10% depending on the length of the initial IO period.
Heh, I just hope that all the 'neg am' income on the books at CFC does come to fruition!
The other biggie looming out there is the option arm recasting. Many of those must be nearing their max LTV considering all that unpaid interest lenders have been uh, not receiving as income. The option arm was especially popular in California.
I would imagine that these are even more difficult to forcast out as it would require even more variables.
presumably these will only be available to a smaller set of borrowers than a year ago. Like, borrowers with actual income proof or say, a downpayment?
Correct. I have attempted to post some estimates on refinance opportunity that take into account both 1) how many borrowers can actually get a lower rate (how many lower rate refis are being made) and 2) how many borrowers can qualify for that lower rate.
The last ballpark number I saw from one of the IBs was that around 36% of currently outstanding subprime ARM borrowers could both qualify for a Fannie/Freddie or FHA loan, and also get a lower rate by doing that.
This estimate will obviously change as either the credit quality of the outstanding loans get better or (likely) worse and rates on new loans get better or worse.
off topic: Freddie Mac was KICKED OUT by American Home Loan security when they tried to get loan files.
"Mortgage Lender's Bankruptcy
May Threaten Thousands of Homeowners
By PEG BRICKLEY Wall Street Journal
September 12, 2007
"Thousands of homeowners face an "imminent risk" of losing their homes because of clashes between American Home Mortgage Investment Corp. and its former financial backers, according to Freddie Mac, a government-chartered housing financier.
In documents filed with the U.S. Bankruptcy Court in Wilmington, Del., Freddie Mac said it seized $7 million that homeowners sent to American Home to cover principal and interest payments, property taxes and insurance just before the company's Aug. 6 collapse. American Home quit making payments to tax authorities and insurance companies Aug. 24.
Freddie Mac said 4,547 loans valued at nearly $797 million are at stake. It said it doesn't have the loan files necessary to pay insurance premiums and property taxes on them, however. "Therefore, there is the imminent risk that borrowers' insurance policies may lapse for nonpayment, subjecting the borrowers to a risk of loss of their mortgaged properties," Freddie Mac said."
[...]
"Just days before American Home's bankruptcy filing, Freddie Mac and Ginnie Mae terminated the company's loan-servicing rights. They also sent representatives to collect loan files from American Home's servicing facility in Irving, Texas.
In court documents, American Home said Ginnie Mae representatives "stood in a line in front of the doors and sat on the stairs, preventing AHM Servicing employees from entering the office." Freddie Mac said American Home "had its security personnel escort the Freddie Mac representatives out."
In addition to Freddie Mac and Ginnie Mae, several Wall Street banks are fighting to extract their loans from American Home's servicing operation. The list includes Morgan Stanley, Deutsche Bank AG, Credit Suisse Group and EMC Mortgage."
Or is the data just too variable to be mapped in this way?
Well, it's not impossible to do that, but it sucks as a project.
You have the IO effect, you have varying terms (30-year, 40-year), you have variable caps and margins and so on. People who do the kind of heavy lifting analytic and number-crunching that this kind of thing requires tend to charge for the results.
Bacon dreamz just gave you one set of estimates. I have another on my desk that puts the payment increase at first adjustment assumption for subprime ARMs at 22%-46% assuming a flat index. (My numbers include both non-IOs and IOs that also amortize at the first adjustment.)
It's not possible to say that one estimate is better or worse, since, as we are discussing, it's a question of what universe of loans you looked at and what you took into account.
Other then stimulating wall street who will a fed funds rate cut help?
We've learned over the past 10 years or so that if the Fed cuts rates it's going to be the speculators and and financial industry that's grown to support them which benefits soonest and mostest.
In the longer-term nobody else really benefits because aggregate wealth production declines as speculators cannibalize productivity for short-term profit.
It's like when researchers found that in giving vitamins to cancer patients, it was the cancer cells that actually benefited most in many cases.
I'm actually optimistic that the Fed "gets this", they just have to overcome political obstacles (e.g. Wall Street lobbyists).
I would imagine that these are even more difficult to forcast out as it would require even more variables.
deb, that really sucks as a project.
The rate of negative amortization on a loan, if you assume the borrower always makes the minimum payment from inception of the loan, is a matter of how deep the teaser rate was and whether there are caps or limits on the periodic rate adjustments.
My ballpark assumption is that at minimum these loans will negatively amortize at 2.5% per year. It can go faster than that, or slower.
You'd have to know the original LTV, the rate of HPA (positive or negative), and the balance cap (ranges from 110% to 125%, I use 115% as an average) to be able to project when they'll hit recast and what their LTV might be like when it happens (which would tell you whether they could refi, realistically, or not).
That problem is much further compounded by the fact that OA is a popular portfolio product. There are securitized OAs, but it's a smaller percent of the OA total than other ARM types.
CR and Tanta I just want to say your blog has better information and maintains a light hearted professionalism not found at any other news source. You guys have created a fabulous community here.
In my opinion this is emerging as the major difficulty in the current market dynamics. An unforseen consequence of syndication. Imagine if CFC ran into some problems. Pure caos
So, a rate cut would NOT fix the housing "problem" (keep housing unaffordable and let people keep houses they could never afford), but WOULD devalue the dollar greatly, leading eventually to fun stuff like $4 a gallon gas, price shock in imports, etc.
Aside from the Wall Street crooks, nobody will benefit. I hope that the Fed and Ben B. realize this and don't cut.
Can you ID your source? FALP? CS? Are you SURE it is current COMBINED CLTV? I mean, I don't know how one either can hope to ascertain (piggybacked or otherwise) seconds behind the OAs, right?
I won't quibble about which HPA index was used, I just don't think they can guesstimate subordinate liens very well. And it's kind of achedemic, anyway. Do you also have current estimated LTV (NOT combined)?
Aaargh...was in the process of the same WSJ AHM link.
I'd like to hear Tanta's thoughts about this. Is this an aberration or is this the future? If so, it seems that the number of "at risk" mortgages has been seriously underestimated.
and I came across the following sentences that I couldn't believe. The article is about subprime resets.
"Many 2/28 hybrid ARM interest rates are based on one-year treasury yields," said said Allen Hardester, a mortgage consultant in Maryland. "The new rate will be more affordable."
From a recent high of 5.02 percent in mid-July, one-year Treasury yields have fallen to 4.09 percent as of Sept. 10. The reset rates of ARMs are calculated using an average of several treasury prices, but the final result should be around that 4.09 percent.
Add a margin of 2.75 percent (a common margin according to Keith Gumbinger of publisher of mortgage information, HSH Associates), and it totals an interest rate of 6.84 percent, compared with 7.77 before."
These are mortgage experts? Every 2/28 loan (which are all subprime) I've ever seen are based on the LIBOR. I don't think I've ever seen one based on the 1 Yr t-bill, certainly not a subprime or alt-A. Conforming and FHA ARM's are based on the 1 Year t-bill. But what about the margin, 2.75% is correct for conforming deals that are tied to the 1 year T-bill, 2.25% for LIBOR based conforming ARM's. For subprime loans, the margin can be anywhere between 5-8%. Someone with a 6% 2/28 subprime loan who has a 6% margin based on the 1 year LIBOR is going to fully reset at 11.75%. Even if the LIBOR went down to 3%, you are still looking at a rate of 9%. There is no way that someone who took out a subprime ARM will be seeing a lower rate, even if they decline. The high margins would never let that happen.
Short version: Fed fund cuts might affect the price of new loans, since it affects the bank's cost of funds and thus its net interest margin. So a rate cut could create refi opportunities. It won't result in rates on outstanding ARMs dropping any time soon.
I think the dropping RE values will prevent refinancing many ARMs. Lower rates are nice, but nobody is going to be too psyched about making a 500k loan with collateral worth only 400k.
Worthy threadjack about Mrs. Zelman. I wonder if there will be much business for her though, it'd be like offering analysis of the dot coms in 2001/2002.
tanta, the unaccounted for additional resets on existing loans is an excellent point and huge. You're right and any forecasts should also include forward looking assumptions, so judgements can be made about what the actual numbers could be.
Models could take this into account and ranges might be useful that consider future LIBOR rates.
I know! Why doun't YOU create some reset charts. You could make a fortune, charging for a peek at the horror show.
Good stuff, as per usual. The one little critique I might offer is that the SECOND reset might not be as innocuous as you suggest, 'cuz of periodic caps restraining the first from it's full potential back-breakitude.
Fed Funds cut to 5.0% will be a non event. Since 8/10 the fed fund rate (based on the official NY Fed statistics) shows the rate is 4.96%. Fed has already lowered. A fed fund rate of 5.0% would actually be a small rise.
What Subsonic22 didn't mention was the title of that CNN/Money article -
"Mortgage reset shock: Not so bad
With a little help from the Fed, borderline borrowers could get some relief from a flood of mortgages whose interest rates are set to jump."
In other words, people with ARMs should bet on the Fed dropping rates, not LIBOR.
In court documents, American Home said Ginnie Mae representatives "stood in a line in front of the doors and sat on the stairs, preventing AHM Servicing employees from entering the office." Freddie Mac said American Home "had its security personnel escort the Freddie Mac representatives out."
So it comes to this.. Sit-ins - LSE (London School of Economics ) style, singing "We shall not be moved" as these bureaucrats get grabbed by their hair and dragged out one by one by security thugs..
Does tear-gassing and truncheon wielding cops follow ?
The one little critique I might offer is that the SECOND reset might not be as innocuous as you suggest
I just meant that the per-adjustment cap is usually 1% on these things. The first adjustment cap is usually at least 2%. So the subsequent adjustments are less wicked from a payment shock perspective because of the cap. I'm certainly not saying that I think LIBOR is going to rally big time.
By Iain Dey, Sunday Telegraph
Last Updated: 12:02am BST 10/09/2007
Britain's biggest banks could be forced to cough up as much as £70bn over the next 10 days, as the credit crisis that has seized the global financial system sparks a fresh wave of chaos.
Seems to me the 11D COFI was pretty prevalent here in California in the 90's. When and why did the shift occur?
Because COFI is a bank/thrift portfolio lender's index. (It is the average costs of deposits and bank borrowings from the Fed District.)
When you get into securitization/secondary market sales, you get investors whose cost of funds is not related to deposits or Fed bank borrowings.
COFI also lags too much for those really evil OAs. They use a moving average Treasury calculation to keep those loans "repriced" to much more current money costs than the good ol' COFI.
15 points on top of start, or 15% administered max?
It doesn't matter whether it's 2% first, 2% annual (1% per) or 3% first, 3% annual, (1.5%) per. You still get the biggest shock at the first adjustment.
Per UBS, WA first cap on the 2/28s is 2.32. WA per cap is 1.17. WA max rate is 14.30 (minus 6.00 is 8.30 WA start rate, which sounds about right to me.) That tells me that the old fashioned last century prime lender delusion 2/1/6 still exists.
Yes, I am often of the vanilla persuasion. But this time I think I'm more chocolate ripple than you think I am.
Yeah, I meant 15 max RATE., not just spread over index. So, we are cool there.
The other two weighted avgs that UBS shows are a wee bit surprising to me, I will be double-chckingthis, choc. ripple chick.
I just don't want faithful readers to draw the conclusion that the mouth-breather who chose a 2/28 for its 'cheaper payment' characteristic is...AH, what the hell,,this thread is nearly closed. Let's talk Horsehockey over some cocktails...
So, I am still confused.
Does anyone know the dollar value or number of resets that are indexed to LIBOR or something other then an index directly affected by the fed funds rate?
Does anyone know the dollar value/number of resets directly affected by the fed funds rate?
My point is ... how will a fed funds rate help lower the reset rate if the reset in not indexed to the a rate that is affected by the fed funds rate?
Now, if the fed funds rate in decrease then the discount rate should follow if it continues to float against the fed funds rate. Is this the real reason for a fed funds rate decrease?
Other then stimulating wall street who will a fed funds rate cut help?
How will a rate cut affect the asset bubble in the housing market?
I could go on and on.
Kudos to you. One of the best pieces on explaining this topic I've seen. An excellent post. Has anyone told you lately this is one of the best blogs out there. Fantastic stuff.
Blogger hates me, this I know
'Cause the page loads are so slow.
Older posts to us belong
But the look-up takes too long.
Yes, Blogger hates me!
Yes, Blogger hates me.
Yes, Blogger hates me,
The page loads tell me so.
Just a little hymn-singing by a frustrated blogger this morning. Hope all you blogees are having a better time than I am.
For those who think a Fed rate cut will help I believe their general theory is the rate cut will naturally draw down all other rate indices. I'm not sold on that theory but it's out there nevertheless.
Tanta,
Thank you! I had wondered why the BofA chart was so much different than the CS chart from before, and had been meaning to ask this very question. I had assumed that it was just the use of proprietary info, but this makes much more sense. So it all depends on what question you ask and what you factor in, which makes using those numbers (for us outsiders, at least) a very inexact tool.
Libor hits high on cash rush
By Joanna Chung
The cost of funds in the hard-hit interbank money markets hit peaks not seen for nearly a decade on Tuesday as the scramble for cash by financial groups showed little sign of easing.
An important borrowing benchmark for investors, the three-month London interbank offered rate (Libor), climbed to 6.90375 per cent in the sterling market, up from 6.89625 on Monday, leaving it 115 basis points above the Bank of Englands base rate of 5.75 per cent and at the highest level seen since November 1998.
Anyone have a link to the Sinai article about many defaults happening even before the first reset?
Sorry, I meant the Ed Hyman report on defaults before reset.
Does anyone know the dollar value or number of resets that are indexed to LIBOR or something other then an index directly affected by the fed funds rate?
Does anyone know the dollar value/number of resets directly affected by the fed funds rate?
I don't have the exact LIBOR percent at my fingertips (and I won't until Blogger stops hating me), but suffice it to say that almost all subprime ARMs are LIBOR-indexed, most Alt-A ARMs are LIBOR, and some prime ARMs are. If a loan isn't LIBOR-indexed, it is most likely to be treasury-indexed. (There is a small pile of weird indexes based on bank cost of funds, or COFI, but ignore that, it's a very small pile these days.)
A whole lot of Option ARMs are treasury-indexed (MTA, or monthly/moving treasury average).
You cannot get directly from Fed funds to either the cost of London interbank borrowing (LIBOR) or U.S. Treasury securities of a constant maturity of one year, six months, or one month, as the case may be for different loan types.
You also must bear in mind that ARMs reset on the schedule in the note, which will depend on the loan's origination date. ARMs that reset October 1 already have their index values determined (from 8/15 or 9/1, depending).
I did a post on that subject in the end of August, and if I could get the forking archive to load properly, I'd link to it.
Short version: Fed fund cuts might affect the price of new loans, since it affects the bank's cost of funds and thus its net interest margin. So a rate cut could create refi opportunities. It won't result in rates on outstanding ARMs dropping any time soon.
This about says it all! Perfect.
http://www.whattolearn.com/FIREWALKER.GIF
Is it feasible to create a graph showing the average (or median) monthly payment for ARMs (all or just securitized). That way.. One could just make assumptions on a range for interest rates and then graph the potential increase in the avg/median monthly payment. You could also just graph the total amount of money payed towards these mortgages over the next 12 months.
Or is the data just too variable to be mapped in this way?
It just seems like this would give a more meaningful illustration of what could really happen dollar wise.
For those who were wondering what Ivy Zelman was up to -
she's started her own research firm and taken a few Credit Suisse people along with her:
Expert on Housing Has Her Own Nest - WSJ.com
Tanta, thank you that was very helpful. Re your comment that a rate cut might create refi opportunities, presumably these will only be available to a smaller set of borrowers than a year ago. Like, borrowers with actual income proof or say, a downpayment?
Eli, the paper on my desk tells me that subprime borrowers facing first reset in the next year ($27B/mo on avg) will get an average payment shock of 35%-45%, based on rate levels on aug. 17. The avg original loan size is like $225K.
and 50% of them have combined LTV over 90%...
Nice piece! I think another whammy has to do with the resets of IO's. Not only do the rates adjust at the end of the fixed rate period, but the payment is recast from an IO to one of a fully amortized loan. This can kick up the monthly payments by 10% depending on the length of the initial IO period.
Heh, I just hope that all the 'neg am' income on the books at CFC does come to fruition!
The other biggie looming out there is the option arm recasting. Many of those must be nearing their max LTV considering all that unpaid interest lenders have been uh, not receiving as income. The option arm was especially popular in California.
I would imagine that these are even more difficult to forcast out as it would require even more variables.
presumably these will only be available to a smaller set of borrowers than a year ago. Like, borrowers with actual income proof or say, a downpayment?
Correct. I have attempted to post some estimates on refinance opportunity that take into account both 1) how many borrowers can actually get a lower rate (how many lower rate refis are being made) and 2) how many borrowers can qualify for that lower rate.
The last ballpark number I saw from one of the IBs was that around 36% of currently outstanding subprime ARM borrowers could both qualify for a Fannie/Freddie or FHA loan, and also get a lower rate by doing that.
This estimate will obviously change as either the credit quality of the outstanding loans get better or (likely) worse and rates on new loans get better or worse.
Tanta,
Another Home Run. Well said. Thanks!
off topic: Freddie Mac was KICKED OUT by American Home Loan security when they tried to get loan files.
"Mortgage Lender's Bankruptcy
May Threaten Thousands of Homeowners
By PEG BRICKLEY Wall Street Journal
September 12, 2007
"Thousands of homeowners face an "imminent risk" of losing their homes because of clashes between American Home Mortgage Investment Corp. and its former financial backers, according to Freddie Mac, a government-chartered housing financier.
In documents filed with the U.S. Bankruptcy Court in Wilmington, Del., Freddie Mac said it seized $7 million that homeowners sent to American Home to cover principal and interest payments, property taxes and insurance just before the company's Aug. 6 collapse. American Home quit making payments to tax authorities and insurance companies Aug. 24.
Freddie Mac said 4,547 loans valued at nearly $797 million are at stake. It said it doesn't have the loan files necessary to pay insurance premiums and property taxes on them, however. "Therefore, there is the imminent risk that borrowers' insurance policies may lapse for nonpayment, subjecting the borrowers to a risk of loss of their mortgaged properties," Freddie Mac said."
[...]
"Just days before American Home's bankruptcy filing, Freddie Mac and Ginnie Mae terminated the company's loan-servicing rights. They also sent representatives to collect loan files from American Home's servicing facility in Irving, Texas.
In court documents, American Home said Ginnie Mae representatives "stood in a line in front of the doors and sat on the stairs, preventing AHM Servicing employees from entering the office." Freddie Mac said American Home "had its security personnel escort the Freddie Mac representatives out."
In addition to Freddie Mac and Ginnie Mae, several Wall Street banks are fighting to extract their loans from American Home's servicing operation. The list includes Morgan Stanley, Deutsche Bank AG, Credit Suisse Group and EMC Mortgage."
Or is the data just too variable to be mapped in this way?
Well, it's not impossible to do that, but it sucks as a project.
You have the IO effect, you have varying terms (30-year, 40-year), you have variable caps and margins and so on. People who do the kind of heavy lifting analytic and number-crunching that this kind of thing requires tend to charge for the results.
Bacon dreamz just gave you one set of estimates. I have another on my desk that puts the payment increase at first adjustment assumption for subprime ARMs at 22%-46% assuming a flat index. (My numbers include both non-IOs and IOs that also amortize at the first adjustment.)
It's not possible to say that one estimate is better or worse, since, as we are discussing, it's a question of what universe of loans you looked at and what you took into account.
Other then stimulating wall street who will a fed funds rate cut help?
We've learned over the past 10 years or so that if the Fed cuts rates it's going to be the speculators and and financial industry that's grown to support them which benefits soonest and mostest.
In the longer-term nobody else really benefits because aggregate wealth production declines as speculators cannibalize productivity for short-term profit.
It's like when researchers found that in giving vitamins to cancer patients, it was the cancer cells that actually benefited most in many cases.
I'm actually optimistic that the Fed "gets this", they just have to overcome political obstacles (e.g. Wall Street lobbyists).
Ready, Set, Reset
The starter holds his gun in the air and the borrowers settle into the blocks...
Bang
And they're off!...oh wait, 25% of them defaulted, they're out of the race. What a shame, they just couldn't handle the pressure of the moment...
I would imagine that these are even more difficult to forcast out as it would require even more variables.
deb, that really sucks as a project.
The rate of negative amortization on a loan, if you assume the borrower always makes the minimum payment from inception of the loan, is a matter of how deep the teaser rate was and whether there are caps or limits on the periodic rate adjustments.
My ballpark assumption is that at minimum these loans will negatively amortize at 2.5% per year. It can go faster than that, or slower.
You'd have to know the original LTV, the rate of HPA (positive or negative), and the balance cap (ranges from 110% to 125%, I use 115% as an average) to be able to project when they'll hit recast and what their LTV might be like when it happens (which would tell you whether they could refi, realistically, or not).
That problem is much further compounded by the fact that OA is a popular portfolio product. There are securitized OAs, but it's a smaller percent of the OA total than other ARM types.
Tanta,
Thanks for the reply.
It is as I suspected, a rate cut will do very little to help with ARM resets.
Thus, it will not prevent most of the possible foreclosures.
and yet another paper on my desk estimates the current combined LTV (including negative amortization and HPA) for securitized OAs as follows:
2004: 59.5
2005: 72.9
2006: 81.2
bacon dreamz, is that origination vintage or change in outstandings by year?
though there is a large number over 90%...the two peaks in the CLTV distribution are at 81% and 92%.
origination vintage.
oops, the 81% and 92% are for the 2006 vintage of course.
I might have voiced optimisim about the Fed too soon. Maybe it's going to take $120/bbl oil to shut these guys up:
Paulson calls for expansion of mortgage products
CR and Tanta I just want to say your blog has better information and maintains a light hearted professionalism not found at any other news source. You guys have created a fabulous community here.
Thank You
rcryan - thanks for posting the update on Ivy Z, even if it was a bit of a threadjack.
In my opinion this is emerging as the major difficulty in the current market dynamics. An unforseen consequence of syndication. Imagine if CFC ran into some problems. Pure caos
Mortgage Lender's Bankruptcy May Threaten Thousands of Homeowners - WSJ.com
So, a rate cut would NOT fix the housing "problem" (keep housing unaffordable and let people keep houses they could never afford), but WOULD devalue the dollar greatly, leading eventually to fun stuff like $4 a gallon gas, price shock in imports, etc.
Aside from the Wall Street crooks, nobody will benefit. I hope that the Fed and Ben B. realize this and don't cut.
Bacon -
Can you ID your source? FALP? CS? Are you SURE it is current COMBINED CLTV? I mean, I don't know how one either can hope to ascertain (piggybacked or otherwise) seconds behind the OAs, right?
I won't quibble about which HPA index was used, I just don't think they can guesstimate subordinate liens very well. And it's kind of achedemic, anyway. Do you also have current estimated LTV (NOT combined)?
Aaargh...was in the process of the same WSJ AHM link.
I'd like to hear Tanta's thoughts about this. Is this an aberration or is this the future? If so, it seems that the number of "at risk" mortgages has been seriously underestimated.
I read the following article today,
Mortgage reset shock: Not so bad - Sep. 12, 2007
and I came across the following sentences that I couldn't believe. The article is about subprime resets.
"Many 2/28 hybrid ARM interest rates are based on one-year treasury yields," said said Allen Hardester, a mortgage consultant in Maryland. "The new rate will be more affordable."
From a recent high of 5.02 percent in mid-July, one-year Treasury yields have fallen to 4.09 percent as of Sept. 10. The reset rates of ARMs are calculated using an average of several treasury prices, but the final result should be around that 4.09 percent.
Add a margin of 2.75 percent (a common margin according to Keith Gumbinger of publisher of mortgage information, HSH Associates), and it totals an interest rate of 6.84 percent, compared with 7.77 before."
These are mortgage experts? Every 2/28 loan (which are all subprime) I've ever seen are based on the LIBOR. I don't think I've ever seen one based on the 1 Yr t-bill, certainly not a subprime or alt-A. Conforming and FHA ARM's are based on the 1 Year t-bill. But what about the margin, 2.75% is correct for conforming deals that are tied to the 1 year T-bill, 2.25% for LIBOR based conforming ARM's. For subprime loans, the margin can be anywhere between 5-8%. Someone with a 6% 2/28 subprime loan who has a 6% margin based on the 1 year LIBOR is going to fully reset at 11.75%. Even if the LIBOR went down to 3%, you are still looking at a rate of 9%. There is no way that someone who took out a subprime ARM will be seeing a lower rate, even if they decline. The high margins would never let that happen.
Shnaps Parlor, i am positive it's combined, it's from RBS. i don't have non-combined LTV ests.
Subsonic22, that's the kind of shit that makes me the crazy person I am today.
Short version: Fed fund cuts might affect the price of new loans, since it affects the bank's cost of funds and thus its net interest margin. So a rate cut could create refi opportunities. It won't result in rates on outstanding ARMs dropping any time soon.
I think the dropping RE values will prevent refinancing many ARMs. Lower rates are nice, but nobody is going to be too psyched about making a 500k loan with collateral worth only 400k.
Worthy threadjack about Mrs. Zelman. I wonder if there will be much business for her though, it'd be like offering analysis of the dot coms in 2001/2002.
tanta, the unaccounted for additional resets on existing loans is an excellent point and huge. You're right and any forecasts should also include forward looking assumptions, so judgements can be made about what the actual numbers could be.
Models could take this into account and ranges might be useful that consider future LIBOR rates.
I know! Why doun't YOU create some reset charts. You could make a fortune, charging for a peek at the horror show.
Tanta-
Good stuff, as per usual. The one little critique I might offer is that the SECOND reset might not be as innocuous as you suggest, 'cuz of periodic caps restraining the first from it's full potential back-breakitude.
Ponder this chart from Dan Green's blog for a sec:
Why LIBOR Will Not Impact Your Adjustable Rate Mortgage This Year | The Mortgage Reports
Fed Funds cut to 5.0% will be a non event. Since 8/10 the fed fund rate (based on the official NY Fed statistics) shows the rate is 4.96%. Fed has already lowered. A fed fund rate of 5.0% would actually be a small rise.
What Subsonic22 didn't mention was the title of that CNN/Money article -
"Mortgage reset shock: Not so bad
With a little help from the Fed, borderline borrowers could get some relief from a flood of mortgages whose interest rates are set to jump."
In other words, people with ARMs should bet on the Fed dropping rates, not LIBOR.
re kip
In court documents, American Home said Ginnie Mae representatives "stood in a line in front of the doors and sat on the stairs, preventing AHM Servicing employees from entering the office." Freddie Mac said American Home "had its security personnel escort the Freddie Mac representatives out."
So it comes to this.. Sit-ins - LSE (London School of Economics ) style, singing "We shall not be moved" as these bureaucrats get grabbed by their hair and dragged out one by one by security thugs..
Does tear-gassing and truncheon wielding cops follow ?
Too funny. Thanks for the laugh..
-K
The one little critique I might offer is that the SECOND reset might not be as innocuous as you suggest
I just meant that the per-adjustment cap is usually 1% on these things. The first adjustment cap is usually at least 2%. So the subsequent adjustments are less wicked from a payment shock perspective because of the cap. I'm certainly not saying that I think LIBOR is going to rally big time.
OT: The second shoe to drop.
\t
Banks face 10-day debt timebomb
By Iain Dey, Sunday Telegraph
Last Updated: 12:02am BST 10/09/2007
Britain's biggest banks could be forced to cough up as much as £70bn over the next 10 days, as the credit crisis that has seized the global financial system sparks a fresh wave of chaos.
http://www.telegraph.co.uk/money/main.jhtml?xml=/money/2007/09/09/cndebt109.xml
ABCP will also reset state side in the few weeks, I heard on CNBC that it was in the neighborhood of $500bl.
So much debt. How much real value?
Tanta,
Seems to me the 11D COFI was pretty prevalent here in California in the 90's. When and why did the shift occur?
Seems to me the 11D COFI was pretty prevalent here in California in the 90's. When and why did the shift occur?
Because COFI is a bank/thrift portfolio lender's index. (It is the average costs of deposits and bank borrowings from the Fed District.)
When you get into securitization/secondary market sales, you get investors whose cost of funds is not related to deposits or Fed bank borrowings.
COFI also lags too much for those really evil OAs. They use a moving average Treasury calculation to keep those loans "repriced" to much more current money costs than the good ol' COFI.
I'm certainly not saying that I think LIBOR is going to rally big time.
why not???
I mean, I'm not predicting that LIBOR could rally enough fast enough to bail out these loans. That would take Mother of Rallies.
Tanta -
I didn't think you were prognosticating LIBOR...'scuse me...'Libor'...rates.
I now see what you were doing there - you think these are 2/1/6 (1st/per/max) cap structures...how quaint. How 5/1 vanilla ARM of you.
Try more like 1.5/15 (per/max).
Now go back and check that graph. Picture LIBOR mostly flat going out a bit. You dig?
Try more like 1.5/15 (per/max).
15 points on top of start, or 15% administered max?
It doesn't matter whether it's 2% first, 2% annual (1% per) or 3% first, 3% annual, (1.5%) per. You still get the biggest shock at the first adjustment.
Per UBS, WA first cap on the 2/28s is 2.32. WA per cap is 1.17. WA max rate is 14.30 (minus 6.00 is 8.30 WA start rate, which sounds about right to me.) That tells me that the old fashioned last century prime lender delusion 2/1/6 still exists.
Yes, I am often of the vanilla persuasion. But this time I think I'm more chocolate ripple than you think I am.
Yeah, I meant 15 max RATE., not just spread over index. So, we are cool there.
The other two weighted avgs that UBS shows are a wee bit surprising to me, I will be double-chckingthis, choc. ripple chick.
I just don't want faithful readers to draw the conclusion that the mouth-breather who chose a 2/28 for its 'cheaper payment' characteristic is...AH, what the hell,,this thread is nearly closed. Let's talk Horsehockey over some cocktails...
Please put this in the Compleat Ubernerd, so I won't need to search for it later.
Thank you!