|Shares Out. (in M):||0||P/E|
|Market Cap (in $M):||1,220||P/FCF|
|Net Debt (in $M):||0||EBIT||0||0|
American Home Mortgage Investment Corporation is the proverbial baby that got thrown out with the subprime bath water. AHM is a REIT that originates and invests in residential mortgages. Subprime loans are just 0.3% of the portfolio and Alt-A loans are 18.2%. Even so, the combined portfolio has a weighted average FICO score of 710 and 75% weighted average loan-to-value (
The question of the moment is liquidity. Unlike its peers, less than 10% of AHM’s funding comes from warehouse credit facilities. They have been able to use purchase and sale agreements, collateralized commercial paper, and collateralized debt obligations (
Here is a summary of their borrowings:
$3.3 billion SLN commercial paper program
$2.0 billion pre-purchase facility with
$2.5 billion facility with Bear Stearns
$1.3 billion bank syndicated facility led by Bank of America, N.A. (committed) *
$125 million facility with J.P. Morgan Chase
$750 million facility with IXIS (committed)
$350 million facility with Credit Suisse First Boston Mortgage Capital LLC
$1.0 billion facility with Barclays
$1.5 billion syndicated facility led by Calyon (committed)
Additional comments from the 10-K
“In addition, we have gestation facilities with
Note: * includes a $446.3 million term loan facility to finance MSRs.
AHM has better credit facilities, but more importantly, they have better loans. The market for prime loans is active and still profitable. The subprime lenders ran into problems when the price of selling subprime loans collapsed. The securitizers stopped paying a premium for these loans. That in turn caused the warehouse lenders, who were often the securitizers, to raise the margin requirement for the warehouse credit facilities. The result was an institutional version of a run on the bank, and the mortgage originators did not have the cash to ante up. If the market for prime loans collapses, there will be a lot of bad things happening.
AHM operates 550 retail and wholesale loan offices in 47 states, which originates 35% of their loan production, with 55% coming from mortgage brokers and 10% from correspondent/direct. By comparison, LEND and
The company recently released an 8-K that provides great information on the loans in their portfolio, broken out by loans held for investment, loans held for sale and securitized loans (which are off-balance sheet):
I like that they have relatively decent FICO scores, low
Management’s guidance for 2007 is loan production of $68 to $74 billion, fully diluted earnings per share of $5.40 to $5.70, and they upped the next quarterly dividend to $1.12/sh. This includes provisioning for more delinquencies.
CEO Michael Strauss owns 9.2% and the CFO has purchased 4,000 shares in the open market the past two weeks.
- cheap at 1.07X book and 1.21X TBV
- nice dividend
- most of the loans and securities they hold are adjustable rate, and are matched to similar liabilities
- while having more stable funding, they do depend upon the kindness of strangers in the capital markets
- mortgage origination business is volatile
- no guarantee that the margins on loan sales remains profitable
- a large part of the loans they hold have stated income
- they are a relatively new company and have only been a REIT since 2004
- 50% of loan portfolio is interest-only
|Entry||03/15/2007 10:22 AM|
|helps lighten my load a little...re the pay-option arms, do you know the duration of their pay-option portfolio? My understanding is that most pay-option borrowers refinance or pay off the loan|
|Entry||03/15/2007 10:58 AM|
The prepayment speed is around 30%, so little over three years. As mentioned in the conference call, AHM tends to see a spike in prepayments shortly after the prepayment penalty period expires. That also explains why 60% of the mortgages they originated last quarter were refinancings.
|Entry||03/15/2007 05:09 PM|
Thanks for the idea.
A few questions:
- what are your thoughts about the Alt-A market? are AHM's originations more skewed towards the Alt-A loans, or are they purchasing them? how much risk do you think is in the Alt-A portfolio?
- how do you think about this as a Reit, versus some of the other mortgage companies that are not Reits?
- how sustainable are the earnings and thus the dividends? it seems to me that as well-capitalized companies like CFC pull back from the sub-prime market, there could be more capacity applied to the prime market, and securitization gains might shrink. do you see that as a meaningful risk? if part of their profits are coming from selling loans they do not originate to companies like CFC, how is that part of the business a sustainable model?
Finally, what do you think of CFC as a buy?
Thank you in advance for your answers.
|Entry||03/15/2007 07:09 PM|
Thanks for the questions. Let me start by reiterating AHM's latest 8-K as a must read because it is full of details on loan production:
I do not have a lot of insight into the Alt-A market, but my thoughts are that a lot of the subprime capacity was outright destroyed and that many originators relying on warehouse credit lines have seen their capacity crimped as capital has to be shifted to support margin calls. I liken the current situation to the position insurers enjoy after a catastrophic event hits and wipes out capital. The survivors will do well.
As for Alt-A loans, they represented 8.9% of total 2006 production for AHM and they represent 8.9% of loans held for investment. 60% of the Alt-A's were put into off-balance sheet securitizations. Total Alt-A's on the balance sheet represent less than 4% of total assets. I am not overly concerned.
On REIT vs. C-corp, the company started out as a C-corp and adopted the REIT structure for tax reasons.
As for the sustainability of earnings & dividends, no one has a good answer. My thoughts are that a rate cut from the Fed will induce more refi's. I also think they have an opportunity to expand market share here.
The subprime market is interesting because CFC may be in a position to expand capacity there. As I said, capacity was destroyed and any originator with capital to hold well underwritten subprime loans is going to be rewarded. Plus, CFC is looking to diversify its business, as evidenced by the switch a thrift charter. It looks like they want to expand into the more lucrative commercial lending area.
I like CFC, and there is a price where I would be a buyer.
|Subject||Insider buying recently|
|Entry||03/16/2007 12:37 PM|
|Michael McManus (Director '01) purchased 5,000 shares at $30.75 on 2/12/07, and an additional 3,000 shares on 3/14/07 at $22.11. Stephen Hozie (CFO '02) purchased 3,000 shares at $25.67 on 3/5/07, and Irving Thau (Director '04) purchased 1,000 shares at $26.05 on 3/6/07. The CFO's purchase is the first time he has bought since 2002. This buying by insiders is the largest in some years.|
|Entry||03/16/2007 05:51 PM|
|Thanks for the write-up. I agree with you on funding sources and minimal exposure to subprime + ALT-A. However, I fret about option ARMs. |
You use the recent 8-k for support. AHM is known for high confidence in its skills & strategies. Case in point: they decided to provide that 8-k intra-quarter. But the tables don’t break down stated/documented income/assets. They also claimed not to have it during the quarterly cc. Furthermore, some data on their website has a bunch of N/As. Even New Century had more detailed data.
Now … on the nature of exotic loan borrowers: There’s a small part of AHM’s 8-k which looks as slick as the rest but probably isn’t: all the rows labeled “Pay-option ARMS”. Here’s what we know:
- BusinessWeek’s Map of Misery demonstrates geographical correlation between Option ARMs density and appreciation rates
- Anecdotal evidence demonstrates similar correlation with incidence of fraud
- Anecdotal evidence: correlation with pressure to “keep up with the Joneses”
- Statistical evidence: correlation with %job growth in real estate sectors (or derivates)
- In other words: option ARM density = multiple risk factors compounded
So let’s say we have two loans A & B with the exact same boring characteristics (say, 720 FICO, 75% CLTV). Loan A is in Vermont and B in Cali. We all agree that - all else being equal - the Cali loan is riskier. Furthermore, let’s consider three scenarios (i) Conventional FRM (ii) typical amortizing ARM (iii) Option ARM. We all agree that as we go from (i) to (ii) and to (iii), the risk spread between A & B increases. Perhaps we even agree that the increase from (ii) to (iii) is much greater than from (i) to (ii). However, I believe financial markets are underestimating the true spread between A and B under scenario (iii).
These risk correlations in my view are based on the feedback mechanism between exotic loans & appreciation on the way up. It’s not just cause + consequence (i.e. exotic mortgage causes appreciation) but rather a loop (loan availability begets speculation, which begets loan availability, which begets…). Then in 2006, option ARMs gained a new role as desperate refinancing tools. And there will be feedback on the way down.
Back to the 8-k: FICOs & LTVs for options ARMs in bubble areas are not meaningful to me. Neither are DTIs.
1) In terms of LTVs, prices rose way too much
2) in terms of DTI, the “I” comes from real estate (reference: Chris Thornburg @ UCLA Anderson)
3) In terms of FICO, if you take a bunch of well-off people and let them speculate and pay minimal monthly nut while being employed by the real estate boom obviously delinquencies will be low! As a matter of fact, this is corroborated by the fact that delinquencies today are highest in the Midwest and South, but not yet in CA, FL, NV, AZ. It is also corroborated by the fact that only subprime has imploded but not the rest, and also by the fact that we have only a “manufacturing recession” but not general recession. A recent survey revealed that 20% of Americans with income 100k+ live paycheck-to-paycheck. I wonder how they are geographically distributed... so average Option ARM borrower FICO can get creamed quite rapidly.
AHM’s 10-k says 23% of loans (excluding HFS) are in CA and 10% FL – not the end of the world but still too concentrated for my taste. If you were to take that 33% and re-spread it in Vermont, Wyoming, West Virginia and Kentucky, maybe I’d be interested.
On top of geographical risk we have the fixed cost of retail origination, the refusal to break-down documentation levels for each loan type, a 20% figure for MI on option ARMs in securitizations and 62% for HFI (which is ok but not super). I’m feeling a bit too chicken to pay stated book value for that. Where am I wrong?
Have you tried contacting the company for the stated/doc statistics?
|Subject||Re: Option ARM Correlations|
|Entry||03/16/2007 09:19 PM|
Let me take a crack at your questions.
> - BusinessWeek’s Map of Misery demonstrates geographical > correlation between Option ARMs density and appreciation > rates
> - Anecdotal evidence demonstrates similar correlation
> with incidence of fraud
> - Anecdotal evidence: correlation with pressure to “keep
> up with the Joneses”
> - Statistical evidence: correlation with %job growth in
> real estate sectors (or derivates)
> - In other words: option ARM density = multiple risk
> factors compounded
For those wishing to see what Carbone is referring to, here is the link to BW's article & Map of Misery:
The biggest issue with BW's article is that they have shown correlation but not causation. Applying anecdotes on top of that makes only for attention grabbing journalism.
> So let’s say we have two loans A & B with the exact same
> boring characteristics (say, 720 FICO, 75% CLTV). Loan A
> is in Vermont and B in Cali. We all agree that - all
> else being equal - the Cali loan is riskier.
No, we don't all agree. You are painting with a broad brush and making underlying assumptions without providing any data points to support them.
> Furthermore, let’s consider three scenarios
> (i) Conventional FRM
> (ii) typical amortizing ARM
> (iii) Option ARM.
> We all agree that as we go from (i) to (ii) and to
> (iii), the risk spread between A & B increases.
Again, not all of us agree. Show me that an option ARM is imperically riskier than a conventional FRM. It may sound like I am playing a semantics game here, but you have to delineate between the product and how it is used. I don't believe that option ARMs are inherently riskier, all things being equal. Are option ARM's riskier in the hands of ethic-challenged mortgage brokers and lenders? Certainly. So far, I have not seen that with AHM.
> Then in 2006, option ARMs gained a new role as desperate
> refinancing tools.
That is an interesting assertion but I think desperation from the originator is more critical than the borrower.
> AHM’s 10-k says 23% of loans (excluding HFS) are in CA
> and 10% FL – not the end of the world but still too
> concentrated for my taste.
Those statistics roughly equate to pro rata population and more closely relate to pro rata home values. Not as bad as LEND which has half its loans for investment in CA.
> If you were to take that 33% and re-spread it in
> Vermont, Wyoming, West Virginia and Kentucky, maybe I’d
> be interested.
Then we'd be dealing with a $50 million market cap!
> On top of geographical risk we have the fixed cost of
> retail origination...
How else do you maintain control?
> ...the refusal to break-down documentation levels for
> each loan type,
> ...a 20% figure for MI on option ARMs in securitizations and 62% for HFI (which is ok but not super).
Mortgage insurance is less critical when LTV < 80%.
> I’m feeling a bit too chicken to pay stated book value
> for that. Where am I wrong?
Everyone has their own risk level.
> Have you tried contacting the company for the stated/doc
Plan on doing that next week. Not specific to AHM but to the morgtage industry in general, how high a hurdle is it to ask a borrower to provide copies of two recent pay stubs?
Having once worked (briefly) for Equifax verifying mortgage applications during the real estate bubble of the late 80's, I know that verifying income from an employer is difficult and time consuming. You had to send a copy of the signed release form from the borrower to an address that they provided. Some employers would come back and ask for an original signed relase form, assuming that it got routed to the right department.
Not condoning it but I suspect that stated income came about because it slowed the process down. For the originator, that meant more warehouse line expense, and mortgage brokers were notorious for directing loans to the lenders with teh quickest yes, which is less of an issue if you have a decent size retail platform like AHM.
|Subject||Why Option ARM's are riskier|
|Entry||03/19/2007 10:59 AM|
|>> Show me that an option ARM is imperically riskier than a |
Assume you want to buy a house. You don't have the income to qualify for a loan, but that's not a problem as you are willing to commit mortgage fraud. Your broker has two mortgages for you to choose from, a fixed rate, and an option ARM. Which product do you choose?
Clearly the option ARM. It offers substantially lower payments to help you squeeze through the first months and years before a default. Whatever scheme you have, an option ARM gives you more time to execute it. Maybe you really think you'll be able to afford the house after you get a new job, a promotion, your dad dies and the will is read, etc, etc.
Option ARMs are much more attractive to mortgage fraudsters, and people who don't have the means to afford their homes. The high payments of fixed loans acts as a "force field" that deflects bad customers, deflecting them over to the Option ARMs.
|Subject||Understanding option ARMs|
|Entry||03/19/2007 12:01 PM|
|The "calculated risk" blog does intense and intelligent coverage (albeit with a very bearish bias) of the whole real estate finance area. One of the contributors, "Tanta", did a long piece the other day explaining in detail how option ARMs work: http://calculatedrisk.blogspot.com/2007/03/tanta-negative-amortization-for.html|
These are very complicated contracts, and it is very rare that either the broker or the borrower completely understand them. One overlooked thing is that there is a limit on how much a borrower can add to the principal balance by choosing to pay less each month than what interest was owed. Once a person has underpaid their interest to the extent that the principal balance has risen 10%-15% (depending on the contract) over the original amount, then the option is taken away, and the loan converts to a conventional one with the borrower required to make full interest payments, plus whatever principal payments it takes to amortize the loan. Many borrowers don't realize this, but after underpaying the interest for long enough, they will get a shocking letter from the lender saying that they have to pay full interest and full principal going forward. Many and possibly most option ARM borrowers won't be able to do that--if they had that kind of dough or income they would have been paying more than the bare minimum all along. This is another bomb that could explode in the next year or two.
|Subject||Re: Option ARM's|
|Entry||03/19/2007 12:22 PM|
Your argument has all the hallmarks of the "She's a witch!" logic from "Monty Python & the Holy Grail."
You are saying that there's plenty of opportunity to commit fraud with options ARMs, and I have already said that I agree on that point, so therefore all option ARM's are bad, which I don't agree with. The problem with such stereotyping is that you have not presented any data to support your thesis. Further, you have not made any attempt to drill down and account for low LTV, higher FICO and the availability of mortgage insurance. Nor have you distinguished loans originated through retail operations vs. those sourced from mortgage brokers.
I don't mind hearing negative stuff about AHM because there is real money on the line here, and one of the benefits of posting an idea at VIC is being able to vet it out. Give me more than rhetoric.
|Entry||03/19/2007 01:58 PM|
Thanks for the lengthy response.
First of all, almost every thesis on housing/mtg stocks lately has ended up in macro discussions because there aren’t any easy-obvious-extreme undervaluations or overvaluations. I have found macro talk to be divisive and good people should not clash about this. Still, I’ll address some of your comments. I hope what I will post is better than bearish rhetoric. I will attempt to once again explain what I mean by “correlation” because I think it’s a huge word in investing.
Fully agree on sensationalist journalism. However, by “anecdotal evidence” I do not mean reading BW or hearing a small number of anecdotes. I’m talking about seeking as much evidence as possible, making historical comparisons, talking to industry sources who are both honest with themselves, competent and experienced, people with neither bearish or bullish agenda. Such people exist and in my view they collectively provide massive evidence that there may be an impending deep unwinding in option-ARM heavy states, partially because of the options ARMs themselves.
It’s ok to play semantics. So let me amend my assertion: “within the context of a frothy speculation-laden market”. [By the way, I am virtually certain that less than 50% of brokers/LOs understand the product, so it’s not just about ethics, it’s a complex product which can only be deciphered by people who sit down and play with the numbers].
Now back to bubble markets: your comment above addresses each loan as an independent unit. But I’m looking at the relationship between units. A loan has many variables: its LTV; the borrower’s DTI; the degree to which the loan will be watched by the FDIC; the conservatism with which the loan will be booked; the way in which the loan will be serviced; the willingness of various parties to refinance the loan; the willingness to buy the loan; to flip the loan; to securitize the loan, the probability of the loan’s owners designating a new servicer…etc. The value of any of these variables in loan A has a interdependency with the value of the same variable in loan B, especially within the same market because real estate is inherently a local market. That’s the correlation I am referring to and that type of dynamic partially generates the causation which BW did not figure out. That causation is multi-directional as you point out and it’s not simply explained. But there is clearly, in my view, something that is happening within Californian society that is not happening within Kentuckian society.
For example, some people (and I don’t necessarily believe them) say that as many as 80% of appraisals in CA/FL were fraudulent. Even if the number is only 50% or 25%, that’s a big enough number for a frothy, competitive housing/brokering/appraising/lending market to generate a mark-to-highest-comp mania. The correlations worked very well on the upside and situations like this eventually end in a similar the downside process. Just look at the recent news: lenders tightening, FDIC roaring, agencies downgrading. Is this the right way to deal with a liquidity crisis? To remove liquidity??? No it’s not; but they can’t help it because they all behaved in a correlated way on the upside.
True in interesting as well. The crack dealer and his customer...
I agree with most of the rest. Interesting anecdote on employment income. On retail, I was being more of a devil’s advocate, it obviously has its pluses and minuses (tie-up capital, vertical integration, operating leverage…).
Looking forward to their response on stated income stuff.
|Entry||03/19/2007 03:04 PM|
|You asked why option arms are empirically riskier than fixed rate mortgages, and I showed you. I did not say every Option Arm is fraudulent or even risky, just that as a class of loans they have higher risk than fixed rate loans because of the type of customers they attract.|
Of course, if a group of option arms are all 30% down, documented income, etc, etc, the risk level is probably very, very low. My original comments weren't directed at AHM, where I have no opinion.
But if you deny that option arms with low credit scores and low down payments are riskier than similar fixed rate loans, it's you who are living in a Pythonesque world.
|Entry||03/19/2007 04:36 PM|
Actually, you did not show me that option ARMs are empirically riskier than fixed rate mortgages. To do that, you would have to track two groups of similar borrowers (FICO, LTV, income, housing market, etc), with one group having options ARMs and the other fixed rate mortgages. That is the only way to show that the mortgage product in and of itself has a different risk profile.
What you are arguing is that option ARMs are attractive to risker borrowers. I agree. Therefore, it is important to look at how a mortgage originator underwrites an option ARM. Since AHM has more skin in the mortgages that they underwrite, both in origination and retention, I believe that they are incented to a better job of originating.
This reminds me of the discussion once about Timberland boots....
|Subject||Interesting Back and Forth...|
|Entry||03/19/2007 09:26 PM|
|...but I'd suggest that a read of an old Golden West Financial Annual Report might help you both settle this debate. Start here:|
My humble opinion is that the contract in and of itself does not compel a certain borrower behavior...therefore a contract can't be bad and those party to it are not destined to deceive or default. Look to the marketing and underwriting.
By the way, Golden West did real well for a real long time. It's now inside of Wachovia. I'm interested in knowing what Wachovia's appetite for option arms is and whether a lack of appetite might explain how this stuff is winding up in new and interesting places (like in NFI's portfolio as well).
|Subject||Problems with Option ARMs|
|Entry||03/19/2007 10:49 PM|
|You can look at the securitizations pools for any of the Alt A originators or go to the sell side and get as many charts as you can imagine which illustrate that the delinquency curve for 2006 Alt A loans is spiking at the same rate as the 2006 subprime delinquency curve. option ARMs as a segment of Alt A loans are able to push delinquencies into the later years given the negative amortization feature so delinquencies there will likely be more back end loaded in this segment of the Alt A universe (but they are still spiking at increasing rates, admittedly off a low base). I find it slightly disturbing that NPAs in AHMs held for sale portfolio went from .43% as of 12/31/05 to 8.13% as of 12/31/06 (from the 10k). And i would also caution that it is very dangerous to look at LTV and FICO averages. I have gone through numersous securitizations in the subprime and ALt A market and all the averages look okay, it is the tails that kill you in the 05 and 06 vintages. I can get the exact stats when i am back in the office, but approximately 9% of AHMs 2006 issuance had between 95-100% CLTVs with low or no documentation and FICO scores under 675 - that seems pretty toxic to me, especially with current pressures on housing prices due to historically high inventory levels. You can start doing the math on delinquencies and severities they can see and the potential write-downs that could occur and keep in mind they are wholly warehouse funded so writedowns could cause their funding base to dissapear pretty fast. On top of all this, cash spreads on Alt A securitizations are widening and that is causing pressure on the Alt A whole loan bid. I think there is alot of risk in AHMs origination platform that you may not be taking into consideration... |
|Entry||03/20/2007 09:07 AM|
Let me respond to some of your concerns.
" I find it slightly disturbing that NPAs in AHMs held for sale portfolio went from .43% as of 12/31/05 to 8.13% as of 12/31/06 (from the 10k)."
Remember that the NPA's are influenced by EPD's from total production that were putted back to AHM, so when looking at defaults as a percentage of all loans, it is not as big. That doesn't lessen the pain that AHM will feel with the loans putted back to them, but if the issue is confined to 2006 production, it is a one-time event.
But let's do some what-if. What if NPA's doubled to 16%. What if loss severity after insurance rose to 25%. That represents a 4% loss of capital. Does it suck? Yes. Does it bankrupt them? No. Does it cut their dividend in half? No.
"I can get the exact stats when i am back in the office, but approximately 9% of AHMs 2006 issuance had between 95-100% CLTVs with low or no documentation and FICO scores under 675 - that seems pretty toxic to me, especially with current pressures on housing prices due to historically high inventory levels. "
I would be interested in seeing those stats. If the figures are total production, then the key is knowing the stats on the loan helds on the balance sheet and in the securitizations. Another important fact would be how much mortgage insurance applied.
I am not saying things are rosy at AHM, but if you take the big scary figures and start running the impact through to AHM, I think there is an over-reaction. If you look at 2006 as a one-off event, that current production is pushing LTV down on lowered home prices, things are going to stabilize.
|Entry||03/20/2007 09:42 AM|
I appreciate the Golden West link. Saved me from misapplying the NRA's old line about "Guns don't kill people, people kill people" to option ARMs.
Bringing up Golden West is also appropriate when looking at how the mortgage landscape has changed in two decades. The intermediation of mortgages has introduced agency risk on the production side, and the new fangled use of warehouse lines and securitizations has heightened liquidity risk. Those poor banking turtles that relied upon cheap deposits and only wrote fixed rate and simple ARM mortgages....
|Entry||03/20/2007 05:45 PM|
|I agree with you that the loan HFS portfolio wont take them down, it is just indicative of the poor underwriting that could take them down. If the NPA on the Loan HFI portfolio keep spiking as they have (up from approximately .31% in 2005 to 1.31% in 2006), the warehouse lenders could require AHM to post additional margin on their credit lines. And these things all feed on themselves as we have seen with NEW. Margin calls make the capital markets even more skittish, which makes the spreads blow out and collateral values decrease which forces more margin calls and the cycle continues.|
My big concern with AHM is a liquidity problem. If AHM makes it through the next 12-24 months without a liquidity problem, you may be right, but i have trouble getting any type of conviction that they will. I guess it just comes down to your probability weighted value for AHM and your probability assignments may differ from mine.
|Entry||05/23/2007 03:34 PM|
|Sorry to interrupt the XS Cargo Network. Here are presentations the company made last week:|
They are the same presentation, with the CFO doing AG Edwards and the CEO doing Lehman. I would listen to the AG Edwards one because the sound is better on the presentation, although they did not mic those asking questions. Sometimes you can hear the questions if you turn up the volume but you can usually get a sense from the response. There were no softball questions. On the Lehman presentation, skip to minute 16, which is about where the Q&A picks up.
When you listen to the presentation, it appears that the blow up in 1st quarter was largely a one-time event due to the high LTV/low doc product that they quickly discontinued, and should largely have played out in 2nd quarter. Most of the delinquencies are early payment default (EPD) and buyers can put the loan back if a EPD
occurs within the first three payments. After that, AHM still provides a warranty on the loan, but it is more about representations as to the loan, i.e., fraud, which is less frequent. Looking at slide #16, 72.8% of delinquencies were related to the discontinued product, which implies that non-accruals should improve in the coming quarters.
Gain on sale margins for the loans are better but not nearly as good. They are still generating enough profit on the loan sales to fund their retail mortgage operations.
They bought a small thrift to generate low cost deposit funding and do some cross-selling.
|Subject||re: various ramblings|
|Entry||05/23/2007 11:29 PM|
|I just re-read this thread and I’ve got some comments on things which were said:|
Option ARMs: Many of us said that option ARMs are a good vehicle for fraud but I think that was a mistake. It doesn’t look like there was much open-eye speculation and open-eye fraud in OAs. What do I mean by “open-eye”? Think of Ben Graham’s distinction between a speculator who knows he’s speculating and one who thinks he’s actually investing. In housing terms: The first guy is the guy who does the EPD, because he knows his attempted flip is royally screwed. The second is someone who bought a home in Cali and spent a little more than he could afford because he “knew prices always go up” and had his MEW. Those are not going to EPD as fast because they actually want to hold on to the home. Similar concept with fraud: The guys who intentionally did a cash-out refinancing so they can get 100k and get the heck out were very comfortable EPDing. However, the guy who stretched his income in a stated income loan because he wanted a nicer house or a bigger house or a second house (or the guy whose broker or LO simply lied without even telling him!) those people will not EPD easily. It’s all about intent. One more fact to remember is that there are not as many 100 LTV OAs as there are amortizing loans (whether stated or not) and open-eye speculators/fraudsters were likely turned off by even a 5% down.
Of course, that is not to say that every OA lender is going to have a smooth ride. The FDIC is sniffing, the secondary market is sniffing. Also, OAs may end up having the worse severities, depending on how you define severity, if there was a 2nd mtg, the MI… And finally, there are a few players who do stated option ARM 100 LTV and that’s just nuts. How many open-eyed people are in these? Open-eye fraudsters are clearly not in them that big. Open-eye speculators likely a bit. One explanation why we’re not seeing worse results in these might be that OA will mega-low teasers are easier to carry than subprime loans with very-low teasers.
David, your latest about how by eliminating 100 CLTV stated (like many did) AHM indeed gets rid of the most troubling part of their biz, is clearly right. I think it’s correct to say that most people out there correctly perceive what loans are deadly, except maybe stated income should be restricted further. The trouble is that AHM still has some crap in their portfolio and I believe the entire peer group will see NPAs slowly but surely flow forward over the next 5 years, and investors will consistently be slightly disappointed with results. That slowness reduces the chance of biv’s liquidity crisis scenario, though it could still happen. Nevertheless, IMO we’re going to see crazy scary levels of inventory and severities. Your thesis is basically right but who knows what hit their BS will take… or what the secondary market will do for them, not only in terms of open/close, but simply margins... so in the end, the etrenal question: who wants AHM to make profit and why?
|Entry||05/24/2007 10:02 AM|
I appreciate the comments. It is not easy buying a falling knife but people tend to over-react in a panic. The difficult part is sorting perceptions from facts, and then trying to determine what the future will be. I read one argument for a person short AHM that as of 3/31/07, NPA's at AHM represented 68% of TBV, whereas that ratio was only 14% as of 12/31/05. One could extrapolate that the next quarter will be horrible.
That completely ignores that the NPA figures from 1st Quarter are distorted by the high LTV/low doc/piggyback loans that they discontinued. Plus, the analysis is missing a 2nd step, of looking at the ratio of actual impairment to NPA's over time.
The catalyst for AHM will be reporting lower NPA's.
|Subject||Q2 Loss Pre-Announcement|
|Entry||06/28/2007 06:38 PM|
|Any thoughts David?|
Managements should never give guidance.
|Entry||06/29/2007 11:36 AM|
|I guess not too many investors read the part in the press release about equity not being imapcted. It is now trading below TBV. I think today's sell-off is largely frustrated people throwing in the towel because they tired of management under delivering. |
|Entry||07/29/2007 12:26 PM|
|What are the mechanics of a declared dividend's delay? A few experienced people have so far said that they've never seen this happen. Anyone know if brokerage/clearance information systems are even equipped to deal with this? |
Let's say someone was short the stock and had to pay the dividend to the securities lender. Can they, like AHM, "decide to delay" the dividend? Will it be arranged automatically? Is it at the discretion of the brokerage?
|Entry||08/02/2007 07:38 PM|
|Well, no fun writing about AHM now, but always important to analyze what went wrong and pay the expensive tuition from the school of experience.|
What did I miss?
I spent so much time looking at the assets that I glanced over the liability side of the balance sheet. Having a dedicated and committed funding source is important for surviving shocks to the financial system. AHM basically borrowed using short-term financing. That works well as long as the acquired assets are liquid. Generally, mortgages and mortgage-backed securities, particularly AAA-rated, are liquid, EXCEPT when there is a major hiccup. The pace at which it occurred has been stunning.
On a side note, a friend of mine made the poignant observation after AHM's collapse that any AAA-rated ABS bond priced at a premium to LIBOR should be really considered a notch below off the bat.
Did not help that their biggest warehouse line was with Bear Stearns....
I have heard positive things about the CEO, Michael Strauss, who owned 4.5 million shares, and yet his silence during all of this is baffling. Losing $100 million of one's worth can certainly cause deer-in-the-headlights syndrome.
|Entry||08/03/2007 10:54 AM|
I also think you chose a company whose management was not of the insincere type. All indications are that they believed in the choices they made and in their company's future, and simply had a classic blow-up based on their lack of prudence.
I prefer Strauss's silence to pompous statements such as "yes there's a downturn out there but we'll be ok, and by the way, U.S. government, you need to help us" that we keeping hearing out of IMB and CFC. Strauss is ashamed and properly behaving in my view.
|Subject||%spent on housing|
|Entry||09/12/2007 11:48 AM|
|Nothing to do with AHM but relevant to our discussion of the nature of borrowers. Another useful map, showing %income spent on housing state-by-state. Kinda correlates with option ARM states (but maybe not causation)|
The point here is that even in (and, on top of that, ESPECIALLY in) these states with a high %option ARMs or I/Os, which arguably should lessen the monthly burden during the non-amortizing period, people are streching. So when option ARMS hit caps and recast...could be hell...