How do you project changes in loan loss when the loan contactually allows the borrower to do negative amortization? If somebody has trouble making increased payments, they just don't!
Well, you see, there is a theory that the very fact that a loan is negatively amortizing implies that the borrower might be in some cash-flow distress.
But that assumption would, taken to its logical conclusion, suggest that one not make neg am loans.
We therefore assume that borrowers are making minimum payments because they are rational informed agents with perfect information who are socking that payment shortfall into some investment that yields more than the interest rate on the loan. Or, perhaps, they are consuming it in ways that keep GDP healthy and thus will prevent a recession, which could be painful because it would increase the number of borrowers making minimum payments for, you know, the bad reasons.
I hope this has cleared everything up. We now return to our regularly-scheduled cognitive dissonance.
Hey - what if they just don't have the extra cabbage?
Obviously this could not be the case, because otherwise:
1) DSL wouldn't be so sure of the 31% they booked in interest earnings that came from the neg-am interest payments (well may be it would have been better had the borrowers paid in cabbage), and
2) they certainly won't have dropped their loan loss accrual to zippo.
Tanta - if you're so smart please explain the following:
1) "The amount of negative amortization included in loan balances increased $37 million during the current quarter to $358 million or 3.56% of loans subject to negative amortization.
2) "approximately 31% of loan interest income represented negative amortization
3) "INTEREST INCOME, Loans $252,172"
4) (my calc) 31% of $252M is $78M - which should be the nominal value of neg-am interest included in the loan balances.
SO, if they report only $37M additional neg am in the loan balances, where is the rest of the $78M?
Could they have rolled the loans over, and administratively cleaned up the spill on aisle 4 - the bad loan aisle?
Well the justification for neg am option loans was always that some people have uneven income streams so that on an annual basis they would have sufficient income to pay an ammortizing amount, but paying the same amount EVERY month might be difficult. All true, all reasonable. That's why they exist. However in recent years they have been used as an "affordability" product to allow people to pay more for a house than they can afford to make monthly payments on. My modest proposal would be to have them recast at 80% CLTV instead of 105%-110% LTV. They are supposed to be alternatives to getthing a seprate HELOC to smooth payments. It's not the occasional period of Negative Ammortization that is the problem with Option Arms. It's the consistant use of them to bring CLTV to unsupportable levels that makes them "suicide loans." Having them recast at 80% CLTV (possibly with regular appraisals) would preserve them for the few cases and whey ARE useful and eliminate much of their negative aspects.
The loans represented by the missing $41M neg-am balances, could these be loans that were approaching the recast which DSL elected to unload?
If so, could DSL claim these as 'seasoned' loans (even though the borrowers were making only minimum payments) while not disclosing the fact that the loans were approaching the recast?
I am reminded of this joke: a guy accidentally fell out of the 100th floor of a building. When he passed the 20th floor, a person in the building leaned out and asked him how he was doing. The answer was "great - so far"
Haps, I suspect they're using net interest income, not gross. Sure, when I do that I get 29%, not 31%, which suggests to me that they really mean net interest on the neg am portfolio, not net interest on the total portfolio.
Jim, I really think you never want an LTV cap on a loan. You want a balance cap (which is what these things have). Who would decide what current value is? How often? Who would pay for all those appraisals? What would this do to you in another bubble? Lenders just have to set the original LTV properly. After that, they have to rely on a balance cap. If you originate at 80% CLTV and you have a 10% balance cap, that means the worst LTV you could get with flat values is 88%.
I still think all that's pointless quibbling when neg am has no place whatsover in the mass market. Nobody should be hockin' the home on a loan they don't understand.
Off track, but reading Commerce Department data, the 44% increase in Midwest housing starts in March is not due to an economic revival-starts in Midwest are 20% below last year at this time and are about par with starts in November and December of 2006. The 44% increase reflects a strong jump from previous February, it probably reflects weather factors more than anything else.
Can anyone shed some light on the "Provision for Credit Losses" line item? Is there some GAAP requirement that states how much Downey needs to set aside for this, or do they just pick a number? Dropping from $10m to just $600k in 1 quarter seems like a desperate stab at making numbers. Since that number is deducted directly from the bottom line it seems more than a bit fishy. Also, the provision / allowance for credit losses only reflects asset impairment, correct? There's nothing to account for the potential booked earnings impact of a neg-am loan that went bad, unless I'm missing something.
The interesting thing about NegArm loans - they do not default. Indeed, how long it will take before they top 110% or 120% - the trigger moment?
I don't think those NegArm banks will ever BK. I think they will report perfect earnings for a while and then Feds will just take over them overnight, one by one.
What I think I see in these numbers (and other stuff in the release) is this:
Total neg am portfolio is down: they're refinancing out and DSL isn't buying/originating more of it at the same rate.
They're getting past their prepayment penalties: note DSL's servicing income stream
Still, the neg am balance growth is huge. 3.56% is a very big number for one quarter.
This suggests to me that DSL's portfolio is "adversely selecting" down to borrowers who always make the minimum payment. Borrowers with refi options (ability to carry amortizing payment) are taking them. So that portfolio not only will keep getting smaller, it will keep getting crappier in terms of concentration of high-risk loans.
Okay, I've triped up on terminology. How about a balance cap of 80% of the orignial appraisal? I think that this would preserve OAs for the few, rare cases where they make sense and almost eliminate the scary upside down moron factor.
Also, the provision / allowance for credit losses only reflects asset impairment, correct? There's nothing to account for the potential booked earnings impact of a neg-am loan that went bad, unless I'm missing something.
Allowance for loan losses is the lender's best estimate, based on its review of the portfolio and its beliefs about the future, of the loan losses it could reasonably incur in the next 12 months. (Capital is there for unanticipated losses.) The reason you adjust reserves every quarter is to keep the 12-month thing.
Another way to think of this is that the allowance lowers the value of the asset on the books by the amount the lender thinks might be uncollectable in the next 12 months.
Neg am deferred interest is added to the loan balance. It is therefore reduced on the books by the loss allowance. That means that the lender is recognizing as interest income only a portion of the neg am balance (net of the allowance).
The problem here, for me at least, is some skepticism about the assumptions DSL put into calculating sufficient reserves. Only God and the regulators can make them change those assumptions. There is no automatic infallible formula for calculating them. It is not a GAAP issue in the sense of how it is presented on the balance sheet/income statement. Under current reserve rules they are allowed to use "historical experience" as a major factor in setting reserves. There is no provision for immediately-previous bubbles that might make immediately-previous historical experience a problematic guide for expectations in the next 12 months.
How does your non-performing assets increase from .32% to .94% (which is a 147% increase YOY)and the allowance for loan loss reserves decrease by $9.4 million?
WASHINGTON (MarketWatch) -- U.S. foreclosure filings increased 7% in March from February's levels and were up 47% from a year ago, according to RealtyTrac, an online real estate database. Nationally, there was one foreclosure filings for every 775 households. Five states -- California, Florida, Texas, Michigan and Ohio -- accounted for half the nation's total in March. In California, foreclosure filings increased 36% from February and were up 183% compared with a year ago. Nevada had the highest foreclosure rate at one in every 183 households, followed by Colorado. Six of the top 10 cities were in California, led by Stockton. The data include default notices, auction sale notices and bank repossessions.
I love the "Neg am during Q1 was $37MM or 3.56% of neg am portfolio, representing 31% of loan interest income " part.
Classic "Unaccountable Accounting" -- set up to book as earnings today the income you claim you're going to get someday because from contracts with people who say they're going to pay you. Then go out and make mountains of contracts without regard to whether the people you're making them with are likely ever actually to pay.
I threw in the towel and covered DSL. It's a crowded short and they aren't going to admit anything is wrong for months. I'll take a look again in early July
Here's an amusing story I heard. A builder in the Phoenix market tore down the model homes and spec houses it build to lower the price point for the development. That's pain in my opinion!
NEW YORK (Reuters) - Washington Mutual Inc. the largest U.S. savings and loan, on Wednesday said it will refinance up to $2 billion of "subprime" home loans at discounted rates to help borrowers who might otherwise struggle to keep up with payments.
Seattle-based WaMu, as the thrift calls itself, said borrowers who are current on existing WaMu subprime loans and anticipate payment increases are eligible to apply for discounted fixed-rate loans and other mortgage products.
I love neg-am booked as current earnings, but the trend I'm most fascinated by now is that despite the flexibility to allow borrowers to make these lower payments, they have doubled the REO assets on their books. Countrywide has a similar growth in their California REO, and I'm assuming many other lenders do as well. From what I've seen of Countrywide's list prices, I'm assuming that the lenders are not discounting the value of these assets yet. That means their balance sheets are even more inflated. Artificially low loan-loss, artificially high REO asset value, artificially inflated interest income...
"... assets totaled $15.238 billion, down $2.565 billion [YOY] ... [neg-am] included in loan balances increased $37 million ... to $358 million or 3.56% of [neg-am loans] ..."
The reciprocal or 3.56% being 28.09, and 28.09 times 358 being 10,056, it looks like two-thirds of their assets are neg-am loans. Am I reading this right?
Stockholder equity as of Dec '06 was $1.4 billion. Operating cash flow in '06 was negative $65 million. Hmmmmm.
The reserve amounts are suspicious, no doubt, but It's feasible that they've already tapped into the reserves to cover losses already realized, and have yet to set reserves based on new information. Besides, the GAAP rules for contingencies and reserves are wide enough to drive a truck through if you document everything, and IAS isn't much better. Basically, to establish a reserve, the event reserved for has to arise out of a past action, the impact has to be measurable, and it has to be More likely than not that the event will occur. So they can say that they think that there is a 49% chance that something will default, and still not reserve for it.
If you originate at 80% CLTV and you have a 10% balance cap, that means the worst LTV you could get with flat values is 88%.
And assuming the CLTV was accurate and the borrower didn't shuffle on down to Ditech 20 minutes after signing. I know a 2nd shouldn't harm the firsts' position but it almost surely recasts the borrower in a lower creditworthiness bin.
Anyone want to take a stab at the ramifications of the Supreme Court ruling yesterday ruling in favor of national banks? On Marketplace (radio), there was a comment that the state authority has been the best to prevent predatory lending, and overall protect consumers.
It is uncommon that one find Stevens and Scalia on the same side, and I wonder if any of our resident bank experts could comment on the lending supervision expectations such a ruling may bring. Tanta?
Also, Second Curve is the first specific hedgie subprime pain I've seen described. This is likely happening elsewhere, and will become more common, I'm sure.
In WaMu's press release yesterday there was no mention of the portion of income that came from neg-am, or even how much neg am (accretion) occurred. They have $67billion in option arms on the balance sheet, and if their accretion rate is anything like CFC's, then nearly 200% of WM's total earnings came from neg am of the Option ARM portfolio at WaMu. This "trick" will work until the re-cast which is probably 6 to 9 months out for most of the loans. FASB is looking at changing the accounting. Will take them another year to get anything done (remember how long it took to do FAS133?).
Another GAAP thought just occured to me. Are the expenses that you accumulate in foreclosing and repossessing a property get booked as expenses, or are they added to the basis you have in the property? I would guess that they are booked as expenses, but I don't know.
Savings associations must initially record foreclosed assets deemed held for sale at the lower of one of the following amounts:
Recorded investment (that is, carrying value before deduction for valuation allowances) in the loan.
Fair value less costs to sell the foreclosed asset.
The costs to sell an asset include the estimated incremental direct costs to transact the sale of the asset. This includes such costs as broker commissions, legal and title transfer fees, and closing costs. Costs to sell generally exclude insurance, security service, and utility costs.
Upon foreclosure (including in-substance foreclosure), the savings association must compare the recorded investment in the loan (carrying value before deduction for valuation allowances) to the fair value less costs to sell the foreclosed property.
The savings association must classify as Loss and charge off any amount in excess of recorded investment over fair value less costs to sell. The savings association cannot represent this Loss classification by a valuation allowance.
Savings associations must expense, as incurred, legal fees and direct costs of acquiring title to foreclosed assets.
I think the question is what is a reasonable reserve in relationship to the non-performing loan. For DSL, the non-perfoming loan is about 0.94%. And the 3/31/07 loss provision: $62m on $12.45b of loans = 0.50%. Anyone has a rule of thumb on what is reasonable reserve? It look o.k. to me since more than 1/2 of non-performing loan normal will get back on payment. And loan loss severity in subprime land is about 30 to 40% of loan once it default. So the acutal loss for 0.94% non-performing loan can be in the range of 0.94x0.5x0.4=0.19% of the overall loan.. Since DSL already reserved 0.5% of loan value. This look o.k. to me. Tanta and other who are more familiar with loan reserve requirement, what is your take?
Non-performing assets increased during the quarter by $33 million to $143 million and represented 0.94% of total assets, compared with 0.68% at year-end 2006 and 0.22% a year ago."
xofruitcake, I'd have two concerns with that. First, look at the rate of increase in NPAs. Reserves are supposed to project out for the next 12 months. You're assuming that 0.94 is the rate over the next year, not just the rate this quarter. But the trendline doesn't justify that to me.
Second, the issue (as the OTS notes) with neg am is that their NPA numbers can be very deceiving: the whole point of neg am is to have a loan that can be in trouble before it appears to be in trouble (i.e., the borrower starts making minimum payments where another borrower with a regularly amortizing loan would just stop making payments). So the rate of increase in capitalized balances must suggest to you that things are not all well for the future. The only thing in my mind that could justify DSL's reserves is the belief that it won't blow up until at least Q207. That doesn't comfort me any.
Downey's negam loans recast at 110% of original balance, unlike First Federal's which recast at 125%. Many of Downey's loans, especially those with margins over 3.0%, which seems to be the majority, will be recasting shortly, in a time frame of 24-36 months from origination, instead of the 5 years that the borrowers originally planned on. The new amortizing payments for the remaining term can be up to triple the current payments, depending on start rate, underlying index and margin.With values declining, the main option open to many borrowers will be to stay in the property, paying the minimum payment, which is a good deal less than what a comparable property would rent for, and then let the property go into foreclosure when it recasts.
Tanta, someone from a different board helped me look up some old DSL filing in the 96. And these were 96 95 94 93 92 91 numbers. In that time frame, their loan loss a a percentage of the loan portfolio is between 1 to 2% and the loss severity is between 45 to 66% of the loan.. So if we go with historic norm (taking the average of these period), the reserve should be around 1.5%x0.55=0.83% of the portfolio. If we think the becuase of neg amt, the loss severity is higher per loan, we are look at may be 1% of the loan portfolio as the correct reserve.. So DSL potentially is under reseved by 62M?? I don't have a position in DSL but I have put on CFC and to me, 62M under reserve is not going to break DSL in anyway. Unless we are making a case to say that the non performing loan will go to 5 or 10% in the next couple quarter, it seems DSL is safe here..
Allowance for loan losses as a percentage of ............. 66.84% 35.67% 49.29% 47.72% 45.30%
non-performing loans
Non-performing assets as a percentage of total assets .... 1.19 2.09 1.41 2.01 1.94
Is it me, or is Downey saying the reason that NegAm declined is because the properties are now REO? The numbers are surpisingly close if I can still do quadratic equations.
Any suggestions for LEAPs? DSL, CORS, and BKUNA do not have LEAPs available. I want to buy some deeply out of the money puts for Jan 2009 and just forget about them.
I decided to short BKUNA, a Florida bank very much like DSL, because I think the FL market is much weaker than even CA. In CA incomes are higher, foreclosure rates are much lower, and new contruction is more limited. LA and SF counties still haven't even seen median price declines yet.
By contrast FL is a nightmare. Builders are still going at almost full speed, incomes are lower, the number of flippers trying to get are are higher, and the big Miami/Palm Beach/FtL markets are down about 10-20% already.
CA will certainly feel pain, but it is about a year behind FL, and will not far quite as hard IMO.
The only thing in my mind that could justify DSL's reserves is the belief that it won't blow up until at least Q207. That doesn't comfort me any.
But how exactly do we view this? I think, fruitcake, your assumption of 1.5% is too conservative. 2% sounds better since this time it's worse. Loss severity should also be worse. And besides, are we truly comparing apples-to-apples? Last time around, there were less OAs, both at DSL and in general. It's a different animal... I think...
Tanta, I agreed with you about apple and orange comparision between DSL now and DSL 10 years ago. However, the key problem that I see is that we have a thread full of people wanting to short all these banks (I am one of it and has some success earlier this year having put on LEND) becuase of the subprime problem and neg amt. But it look to me that there is still a long, long way before DSL or any of those neg Amt guy blow up and has to take hugh reserve. If you believe DSL already taken enough reserve for 0.96% of non-performing loan, having a short position today mean that one is thinking that DSL will have 5% non-performing loan in 3-6 months. And if the trigger is the 110% neg amt cap, we are looking at early next year at the earliest. Put position is going to get kill in between now and then (DSL only has short duration put to Nov for now).
I think the subprime guy using 40% or so for loss severity. So may be DSL will have 50-60% loss becuase of the neg amt. But 5% loan loss with 60% severity will put their reserve at around 1.25Bx0.05x0.6=375M. And DSL stockholder equity is about 1.45B. Don't think that DSL will go under even with this set of number (stock price can be cripple). By the way, BKUNA already report some very good number tonight.
By the way, thanks for hosting all the discussion about mortgage. I learned a great deal from your discussion. Hopefully you make a bunch with all the knowledge that you have in this industry. I can say for sure that someone will make or loss a lot of money in stocks in this industry this year and next.
Yes thanks to Tanta. I am amazed at how little I understand about mortgage accounting, and I suspect I know more than 99.9% of market participants.
But I have always known one thing: with financials, the balance sheet is everything, the income statement very little. Given how hard it is to assess balance sheets from the outside, I have done very well over the years to avoid these industries entirely. [I don't short at all.]
fruitcake. That was me, not Tanta... on to your comment:
But it look to me that there is still a long, long way before DSL or any of those neg Amt guy blow up and has to take hugh reserve.
Well, I too, as a short in some of these stocks, worry that I'm shorting something that's already been shorted enough, has declined enough and that has adequately reserved itself. However, I look at the subprime space and I ask "ok these guys sure as heck were over-optimistic (including even LEND and NFI) and the market treated them according to what the market was fed BY THEM, so why should the BKUNAs FEDs and DSLs of the world be much cleaner? Even AHM, NDE, IMH, CFC are proving that not all is well. So what... DSL's price is discounted for future problems? really? The only time DSL declined was a while ago. The subprime mess did not make it decline wquite that much, since the market's reaction was "this is limited to subprime".
So that's why I take it all witha grain of salt. One area, by the way, that you might wanna take a look at, is the SG&A that needs to be supported. Will these guys lower standards for volume? There are rumors that DSL is doing just that.
How do you project changes in loan loss when the loan contactually allows the borrower to do negative amortization? If somebody has trouble making increased payments, they just don't!
Well, you see, there is a theory that the very fact that a loan is negatively amortizing implies that the borrower might be in some cash-flow distress.
But that assumption would, taken to its logical conclusion, suggest that one not make neg am loans.
We therefore assume that borrowers are making minimum payments because they are rational informed agents with perfect information who are socking that payment shortfall into some investment that yields more than the interest rate on the loan. Or, perhaps, they are consuming it in ways that keep GDP healthy and thus will prevent a recession, which could be painful because it would increase the number of borrowers making minimum payments for, you know, the bad reasons.
I hope this has cleared everything up. We now return to our regularly-scheduled cognitive dissonance.
Hey - what if they just don't have the extra cabbage?
Obviously this could not be the case, because otherwise:
1) DSL wouldn't be so sure of the 31% they booked in interest earnings that came from the neg-am interest payments (well may be it would have been better had the borrowers paid in cabbage), and
2) they certainly won't have dropped their loan loss accrual to zippo.
But hey - it's all good news.
DSL trades at over $64 a share and was up in after-hours trading.
[to those of you who do puts...]
Okay.
Tanta, please describe to me the risk reduction techniques that permit DSL to be traded at all.
They appear to be sitting on a keg of dynamite with the fuse burned down to the bung hole, and the stock market is trading the thing like it was GE.
I need to know more about risk reduction.
Tanta - if you're so smart please explain the following:
1) "The amount of negative amortization included in loan balances increased $37 million during the current quarter to $358 million or 3.56% of loans subject to negative amortization.
2) "approximately 31% of loan interest income represented negative amortization
3) "INTEREST INCOME, Loans $252,172"
4) (my calc) 31% of $252M is $78M - which should be the nominal value of neg-am interest included in the loan balances.
SO, if they report only $37M additional neg am in the loan balances, where is the rest of the $78M?
Could they have rolled the loans over, and administratively cleaned up the spill on aisle 4 - the bad loan aisle?
arbogast-
where do I get a hold of the after-hours trading info?
Well the justification for neg am option loans was always that some people have uneven income streams so that on an annual basis they would have sufficient income to pay an ammortizing amount, but paying the same amount EVERY month might be difficult. All true, all reasonable. That's why they exist. However in recent years they have been used as an "affordability" product to allow people to pay more for a house than they can afford to make monthly payments on. My modest proposal would be to have them recast at 80% CLTV instead of 105%-110% LTV. They are supposed to be alternatives to getthing a seprate HELOC to smooth payments. It's not the occasional period of Negative Ammortization that is the problem with Option Arms. It's the consistant use of them to bring CLTV to unsupportable levels that makes them "suicide loans." Having them recast at 80% CLTV (possibly with regular appraisals) would preserve them for the few cases and whey ARE useful and eliminate much of their negative aspects.
Tanta - follow up question:
The loans represented by the missing $41M neg-am balances, could these be loans that were approaching the recast which DSL elected to unload?
If so, could DSL claim these as 'seasoned' loans (even though the borrowers were making only minimum payments) while not disclosing the fact that the loans were approaching the recast?
I am reminded of this joke: a guy accidentally fell out of the 100th floor of a building. When he passed the 20th floor, a person in the building leaned out and asked him how he was doing. The answer was "great - so far"
Haps, I suspect they're using net interest income, not gross. Sure, when I do that I get 29%, not 31%, which suggests to me that they really mean net interest on the neg am portfolio, not net interest on the total portfolio.
Jim, I really think you never want an LTV cap on a loan. You want a balance cap (which is what these things have). Who would decide what current value is? How often? Who would pay for all those appraisals? What would this do to you in another bubble? Lenders just have to set the original LTV properly. After that, they have to rely on a balance cap. If you originate at 80% CLTV and you have a 10% balance cap, that means the worst LTV you could get with flat values is 88%.
I still think all that's pointless quibbling when neg am has no place whatsover in the mass market. Nobody should be hockin' the home on a loan they don't understand.
Off track, but reading Commerce Department data, the 44% increase in Midwest housing starts in March is not due to an economic revival-starts in Midwest are 20% below last year at this time and are about par with starts in November and December of 2006. The 44% increase reflects a strong jump from previous February, it probably reflects weather factors more than anything else.
Can anyone shed some light on the "Provision for Credit Losses" line item? Is there some GAAP requirement that states how much Downey needs to set aside for this, or do they just pick a number? Dropping from $10m to just $600k in 1 quarter seems like a desperate stab at making numbers. Since that number is deducted directly from the bottom line it seems more than a bit fishy.
Also, the provision / allowance for credit losses only reflects asset impairment, correct? There's nothing to account for the potential booked earnings impact of a neg-am loan that went bad, unless I'm missing something.
The interesting thing about NegArm loans - they do not default. Indeed, how long it will take before they top 110% or 120% - the trigger moment?
I don't think those NegArm banks will ever BK. I think they will report perfect earnings for a while and then Feds will just take over them overnight, one by one.
Although, is it surpising that the weather factor is trotted out to discount "bad" data, but is never mentioned as a factor in "good" data?
What I think I see in these numbers (and other stuff in the release) is this:
Total neg am portfolio is down: they're refinancing out and DSL isn't buying/originating more of it at the same rate.
They're getting past their prepayment penalties: note DSL's servicing income stream
Still, the neg am balance growth is huge. 3.56% is a very big number for one quarter.
This suggests to me that DSL's portfolio is "adversely selecting" down to borrowers who always make the minimum payment. Borrowers with refi options (ability to carry amortizing payment) are taking them. So that portfolio not only will keep getting smaller, it will keep getting crappier in terms of concentration of high-risk loans.
Okay, I've triped up on terminology. How about a balance cap of 80% of the orignial appraisal? I think that this would preserve OAs for the few, rare cases where they make sense and almost eliminate the scary upside down moron factor.
Also, the provision / allowance for credit losses only reflects asset impairment, correct? There's nothing to account for the potential booked earnings impact of a neg-am loan that went bad, unless I'm missing something.
Allowance for loan losses is the lender's best estimate, based on its review of the portfolio and its beliefs about the future, of the loan losses it could reasonably incur in the next 12 months. (Capital is there for unanticipated losses.) The reason you adjust reserves every quarter is to keep the 12-month thing.
Another way to think of this is that the allowance lowers the value of the asset on the books by the amount the lender thinks might be uncollectable in the next 12 months.
Neg am deferred interest is added to the loan balance. It is therefore reduced on the books by the loss allowance. That means that the lender is recognizing as interest income only a portion of the neg am balance (net of the allowance).
The problem here, for me at least, is some skepticism about the assumptions DSL put into calculating sufficient reserves. Only God and the regulators can make them change those assumptions. There is no automatic infallible formula for calculating them. It is not a GAAP issue in the sense of how it is presented on the balance sheet/income statement. Under current reserve rules they are allowed to use "historical experience" as a major factor in setting reserves. There is no provision for immediately-previous bubbles that might make immediately-previous historical experience a problematic guide for expectations in the next 12 months.
Completely cooked the books.
How does your non-performing assets increase from .32% to .94% (which is a 147% increase YOY)and the allowance for loan loss reserves decrease by $9.4 million?
WASHINGTON (MarketWatch) -- U.S. foreclosure filings increased 7% in March from February's levels and were up 47% from a year ago, according to RealtyTrac, an online real estate database. Nationally, there was one foreclosure filings for every 775 households. Five states -- California, Florida, Texas, Michigan and Ohio -- accounted for half the nation's total in March. In California, foreclosure filings increased 36% from February and were up 183% compared with a year ago. Nevada had the highest foreclosure rate at one in every 183 households, followed by Colorado. Six of the top 10 cities were in California, led by Stockton. The data include default notices, auction sale notices and bank repossessions.
Foreclosures up 47% year-on-year in March: RealtyTrac
I love the "Neg am during Q1 was $37MM or 3.56% of neg am portfolio, representing 31% of loan interest income " part.
Classic "Unaccountable Accounting" -- set up to book as earnings today the income you claim you're going to get someday because from contracts with people who say they're going to pay you. Then go out and make mountains of contracts without regard to whether the people you're making them with are likely ever actually to pay.
Jim, if you wanted to cap the borrower at 80% based on original value, you would just make the max original value 72.5% with a 10% balance cap.
Frank, I think there's a difference between "cooking the books" and being stupid. I vote for door number 2 at this point.
I threw in the towel and covered DSL. It's a crowded short and they aren't going to admit anything is wrong for months. I'll take a look again in early July
Here's an amusing story I heard. A builder in the Phoenix market tore down the model homes and spec houses it build to lower the price point for the development. That's pain in my opinion!
NEW YORK (Reuters) - Washington Mutual Inc. the largest U.S. savings and loan, on Wednesday said it will refinance up to $2 billion of "subprime" home loans at discounted rates to help borrowers who might otherwise struggle to keep up with payments.
Seattle-based WaMu, as the thrift calls itself, said borrowers who are current on existing WaMu subprime loans and anticipate payment increases are eligible to apply for discounted fixed-rate loans and other mortgage products.
Washington Mutual commits $2 bln for subprime help
| Reuters
I love neg-am booked as current earnings, but the trend I'm most fascinated by now is that despite the flexibility to allow borrowers to make these lower payments, they have doubled the REO assets on their books. Countrywide has a similar growth in their California REO, and I'm assuming many other lenders do as well. From what I've seen of Countrywide's list prices, I'm assuming that the lenders are not discounting the value of these assets yet. That means their balance sheets are even more inflated. Artificially low loan-loss, artificially high REO asset value, artificially inflated interest income...
"... assets totaled $15.238 billion, down $2.565 billion [YOY] ... [neg-am] included in loan balances increased $37 million ... to $358 million or 3.56% of [neg-am loans] ..."
The reciprocal or 3.56% being 28.09, and 28.09 times 358 being 10,056, it looks like two-thirds of their assets are neg-am loans. Am I reading this right?
Stockholder equity as of Dec '06 was $1.4 billion. Operating cash flow in '06 was negative $65 million. Hmmmmm.
The reserve amounts are suspicious, no doubt, but It's feasible that they've already tapped into the reserves to cover losses already realized, and have yet to set reserves based on new information. Besides, the GAAP rules for contingencies and reserves are wide enough to drive a truck through if you document everything, and IAS isn't much better. Basically, to establish a reserve, the event reserved for has to arise out of a past action, the impact has to be measurable, and it has to be More likely than not that the event will occur. So they can say that they think that there is a 49% chance that something will default, and still not reserve for it.
Shady? Absolutely.
Improper accounting? Hardly.
If you originate at 80% CLTV and you have a 10% balance cap, that means the worst LTV you could get with flat values is 88%.
And assuming the CLTV was accurate and the borrower didn't shuffle on down to Ditech 20 minutes after signing. I know a 2nd shouldn't harm the firsts' position but it almost surely recasts the borrower in a lower creditworthiness bin.
Anyone want to take a stab at the ramifications of the Supreme Court ruling yesterday ruling in favor of national banks? On Marketplace (radio), there was a comment that the state authority has been the best to prevent predatory lending, and overall protect consumers.
Watters v Wachovia
It is uncommon that one find Stevens and Scalia on the same side, and I wonder if any of our resident bank experts could comment on the lending supervision expectations such a ruling may bring. Tanta?
Also, Second Curve is the first specific hedgie subprime pain I've seen described. This is likely happening elsewhere, and will become more common, I'm sure.
In WaMu's press release yesterday there was no mention of the portion of income that came from neg-am, or even how much neg am (accretion) occurred. They have $67billion in option arms on the balance sheet, and if their accretion rate is anything like CFC's, then nearly 200% of WM's total earnings came from neg am of the Option ARM portfolio at WaMu. This "trick" will work until the re-cast which is probably 6 to 9 months out for most of the loans. FASB is looking at changing the accounting. Will take them another year to get anything done (remember how long it took to do FAS133?).
It turns out this:
REOs Stats
was very accurate
Yal,
Where is your source for that spreadsheet?
Another GAAP thought just occured to me. Are the expenses that you accumulate in foreclosing and repossessing a property get booked as expenses, or are they added to the basis you have in the property? I would guess that they are booked as expenses, but I don't know.
incessant_din, thus spake OTS:
Savings associations must initially record foreclosed assets deemed held for sale at the lower of one of the following amounts:
Recorded investment (that is, carrying value before deduction for valuation allowances) in the loan.
Fair value less costs to sell the foreclosed asset.
The costs to sell an asset include the estimated incremental direct costs to transact the sale of the asset. This includes such costs as broker commissions, legal and title transfer fees, and closing costs. Costs to sell generally exclude insurance, security service, and utility costs.
Upon foreclosure (including in-substance foreclosure), the savings association must compare the recorded investment in the loan (carrying value before deduction for valuation allowances) to the fair value less costs to sell the foreclosed property.
The savings association must classify as Loss and charge off any amount in excess of recorded investment over fair value less costs to sell. The savings association cannot represent this Loss classification by a valuation allowance.
Savings associations must expense, as incurred, legal fees and direct costs of acquiring title to foreclosed assets.
http://www.ots.gov/docs/4/422083.pdf
I already have a short position in DSL, but I'm sorely tempted to add to it on today's post-earnings "rally".
This smells like a classic shorting opportunity. For now, however, I'm waiting and watching.
I think the question is what is a reasonable reserve in relationship to the non-performing loan. For DSL, the non-perfoming loan is about 0.94%. And the 3/31/07 loss provision: $62m on $12.45b of loans = 0.50%. Anyone has a rule of thumb on what is reasonable reserve? It look o.k. to me since more than 1/2 of non-performing loan normal will get back on payment. And loan loss severity in subprime land is about 30 to 40% of loan once it default. So the acutal loss for 0.94% non-performing loan can be in the range of 0.94x0.5x0.4=0.19% of the overall loan.. Since DSL already reserved 0.5% of loan value. This look o.k. to me. Tanta and other who are more familiar with loan reserve requirement, what is your take?
Expired? .v=100
"Non-Performing Assets
Non-performing assets increased during the quarter by $33 million to $143 million and represented 0.94% of total assets, compared with 0.68% at year-end 2006 and 0.22% a year ago."
xofruitcake, I'd have two concerns with that. First, look at the rate of increase in NPAs. Reserves are supposed to project out for the next 12 months. You're assuming that 0.94 is the rate over the next year, not just the rate this quarter. But the trendline doesn't justify that to me.
Second, the issue (as the OTS notes) with neg am is that their NPA numbers can be very deceiving: the whole point of neg am is to have a loan that can be in trouble before it appears to be in trouble (i.e., the borrower starts making minimum payments where another borrower with a regularly amortizing loan would just stop making payments). So the rate of increase in capitalized balances must suggest to you that things are not all well for the future. The only thing in my mind that could justify DSL's reserves is the belief that it won't blow up until at least Q207. That doesn't comfort me any.
Downey's negam loans recast at 110% of original balance, unlike First Federal's which recast at 125%. Many of Downey's loans, especially those with margins over 3.0%, which seems to be the majority, will be recasting shortly, in a time frame of 24-36 months from origination, instead of the 5 years that the borrowers originally planned on. The new amortizing payments for the remaining term can be up to triple the current payments, depending on start rate, underlying index and margin.With values declining, the main option open to many borrowers will be to stay in the property, paying the minimum payment, which is a good deal less than what a comparable property would rent for, and then let the property go into foreclosure when it recasts.
Tanta, someone from a different board helped me look up some old DSL filing in the 96. And these were 96 95 94 93 92 91 numbers. In that time frame, their loan loss a a percentage of the loan portfolio is between 1 to 2% and the loss severity is between 45 to 66% of the loan.. So if we go with historic norm (taking the average of these period), the reserve should be around 1.5%x0.55=0.83% of the portfolio. If we think the becuase of neg amt, the loss severity is higher per loan, we are look at may be 1% of the loan portfolio as the correct reserve.. So DSL potentially is under reseved by 62M?? I don't have a position in DSL but I have put on CFC and to me, 62M under reserve is not going to break DSL in anyway. Unless we are making a case to say that the non performing loan will go to 5 or 10% in the next couple quarter, it seems DSL is safe here..
http://edgar.sec.gov/Archives/edgar/data/935063/0000935063-97-000002.txt
Allowance for loan losses as a percentage of ............. 66.84% 35.67% 49.29% 47.72% 45.30%
non-performing loans
Non-performing assets as a percentage of total assets .... 1.19 2.09 1.41 2.01 1.94
Tanta:Nobody should be hockin' the home on a loan they don't understand.
I put together a little visual backup...
http://thumbsnap.com/v/3Tuk4NgH.jpg
Is it me, or is Downey saying the reason that NegAm declined is because the properties are now REO? The numbers are surpisingly close if I can still do quadratic equations.
Looks like 32% of all DSL shares are short as of a month ago... Not a very good sign....
Any suggestions for LEAPs? DSL, CORS, and BKUNA do not have LEAPs available. I want to buy some deeply out of the money puts for Jan 2009 and just forget about them.
How about NDE?
I decided to short BKUNA, a Florida bank very much like DSL, because I think the FL market is much weaker than even CA. In CA incomes are higher, foreclosure rates are much lower, and new contruction is more limited. LA and SF counties still haven't even seen median price declines yet.
By contrast FL is a nightmare. Builders are still going at almost full speed, incomes are lower, the number of flippers trying to get are are higher, and the big Miami/Palm Beach/FtL markets are down about 10-20% already.
CA will certainly feel pain, but it is about a year behind FL, and will not far quite as hard IMO.
The only thing in my mind that could justify DSL's reserves is the belief that it won't blow up until at least Q207. That doesn't comfort me any.
But how exactly do we view this? I think, fruitcake, your assumption of 1.5% is too conservative. 2% sounds better since this time it's worse. Loss severity should also be worse. And besides, are we truly comparing apples-to-apples? Last time around, there were less OAs, both at DSL and in general. It's a different animal... I think...
Tanta?
Tanta, I agreed with you about apple and orange comparision between DSL now and DSL 10 years ago. However, the key problem that I see is that we have a thread full of people wanting to short all these banks (I am one of it and has some success earlier this year having put on LEND) becuase of the subprime problem and neg amt. But it look to me that there is still a long, long way before DSL or any of those neg Amt guy blow up and has to take hugh reserve. If you believe DSL already taken enough reserve for 0.96% of non-performing loan, having a short position today mean that one is thinking that DSL will have 5% non-performing loan in 3-6 months. And if the trigger is the 110% neg amt cap, we are looking at early next year at the earliest. Put position is going to get kill in between now and then (DSL only has short duration put to Nov for now).
I think the subprime guy using 40% or so for loss severity. So may be DSL will have 50-60% loss becuase of the neg amt. But 5% loan loss with 60% severity will put their reserve at around 1.25Bx0.05x0.6=375M. And DSL stockholder equity is about 1.45B. Don't think that DSL will go under even with this set of number (stock price can be cripple). By the way, BKUNA already report some very good number tonight.
By the way, thanks for hosting all the discussion about mortgage. I learned a great deal from your discussion. Hopefully you make a bunch with all the knowledge that you have in this industry. I can say for sure that someone will make or loss a lot of money in stocks in this industry this year and next.
Yes thanks to Tanta. I am amazed at how little I understand about mortgage accounting, and I suspect I know more than 99.9% of market participants.
But I have always known one thing: with financials, the balance sheet is everything, the income statement very little. Given how hard it is to assess balance sheets from the outside, I have done very well over the years to avoid these industries entirely. [I don't short at all.]
DSL update From Amit Chokshi:
Still Short Downey Financial Despite Jump on Earnings -- Seeking Alpha
fruitcake. That was me, not Tanta... on to your comment:
But it look to me that there is still a long, long way before DSL or any of those neg Amt guy blow up and has to take hugh reserve.
Well, I too, as a short in some of these stocks, worry that I'm shorting something that's already been shorted enough, has declined enough and that has adequately reserved itself. However, I look at the subprime space and I ask "ok these guys sure as heck were over-optimistic (including even LEND and NFI) and the market treated them according to what the market was fed BY THEM, so why should the BKUNAs FEDs and DSLs of the world be much cleaner? Even AHM, NDE, IMH, CFC are proving that not all is well. So what... DSL's price is discounted for future problems? really? The only time DSL declined was a while ago. The subprime mess did not make it decline wquite that much, since the market's reaction was "this is limited to subprime".
So that's why I take it all witha grain of salt. One area, by the way, that you might wanna take a look at, is the SG&A that needs to be supported. Will these guys lower standards for volume? There are rumors that DSL is doing just that.