Shopping at Amazon?
Wait! If you click on the Amazon banner link at the top of the page first, HCN will get a percentage of anything you purchase for 24 hours!
Open predictions
Our local book authors
User login
Who's online
There are currently 34 members and 172 guests online.
Currently online
- picosec
- MattFea
- BarleyReturns
- Feckless Ness
- Rajesh
- Bad Dawg Bobby
- Former Idealist
- nova
- josap
- ResistanceIsFeudal
- CalculatedRisk
- bearly
- Tom Stone
- shill
- Comrade Kristina
- aleister perdurabo
- prairiedog
- RayOnTheFarm
- RATM
- Mike in Long Island
- traderwalt
- Spatch
- montas ankle
- butter
- Lazlo
- Blackhalo
- 1 currency now -yogi
- Comrade Alexei Mikhailovich
- adornosghost
- Antipodes
- MSM
- sporkfed
- dr munch
- dingojoe
Amen.
I never understood about this buy back the loan thing. None of the originators I ever worked with had the money to buy back a loan. Mtg brokers sure don't have enough money to buy back even one loan. The mtg brokers I used to work with closed their business because they couldn't place any loans any more. They followed the rules as set out for them. They did do 2-28 mtges, but didn't do any neg am loans (except one for a sophisticated borrower who knew exactly what he was getting into), because they didn't believe in them. Countrywide was always trying to entice them into doing toxic loans, but they would just tell the rep, that for the people who qualified for better, their customers wanted fixed rates.
When they shut their doors, they had business which could have been placed a year ago, but no one would take it now. Some of it didn't sound so bad to me.
I think there is or will be a certain amount of throwing the baby out with the bathwater, expecially after real estate prices go down another 20 or 25%.
Private investers my guys used, who loan at just under usorious limits in Florida, were maxed out. Even people who were paying 17% or 18% couldn't refi, even tho these same investors were using their own money and therefore left themselves a nice equity cushion. (or, so they thought at the time; now who knows what anything is worth?)
The solution is for investors to refuse to get within 20 feet of a mortgage-backed security that is backed by dumb loans. If you do not know what a dumb loan is, you might want to consider investing in a different kind of instrument.
amen. that must have been some good coffee you brewed.
It was great coffee.
Unfortunately, I have apparently said "drop dead" to Verizon DSL too many times, since it obliged me three times this morning by doing exactly that.
This is so, "if Johnny jumps off a cliff are you going to jump off a cliff?"
Seems a disconnect in some adult's ability to actually apply what we try to teach from them since they were toddlers.
Tanta, this is so right, you pinko commie witch.
There's proof, too, at Edgar. Anyone can pull the FWPs on a sample of the really problematic trusts and read. It is hair-raising stuff - underwriters allowed to raise appraisals, for example. First day out of bankruptcy gets 95% LTVs, low-doc. Of course 95% LTV doesn't mean much when your underwriter is raising appraisal values.
There was fraud, but there was also just plain reckless behavior on the part of investors and the aggregators. No one can convince me that they didn't know what they were doing. It was just very profitable for a while.
And now that it's not, they want the US taxpayer to guarantee these?
Tanta,
As always a pleasure to read. I don't have any background in mortgage/RE finance so the work you and CR publish every day helps me follow this fascinating unwinding.
One very small quibble/suggestion - my eyes lit up in the 4th from last paragraph, where you write "The essential confusion here is between failure to follow responsible guidelines and faithful following of irresponsible guidelines. My sad news for the investment community: a whole lot of what you are suffering from is the latter, not the former."
I think in some circles this might be called "burying the lede." Once I'd read that sentence, I think everything in the rest of the piece made sense (I know you wrote above about your blood pressure and ALS).
There's a logic in bubbles that is similar to Parkinson's law, about people being promoted to the level of their incompetence. Some people start making money off a new technique or product, but then the sales targets keep getting stretched and stretched - until we reach the limit of society's incompetence. Irresponsible guidelines are a result. And these irresponsible guidelines seem, at the time, logical and smart "BHAGs" to those issuing them, though not, perhaps, to some veterans who see the pitfalls.
Anyhow, thanks again for the exposition. You two are a bright light in a naughty world.
So, what exactly is the benefit of complicated securitization deals over just having banks issue mortgages with money borrowed via long-term instruments? Aggregation and disaggregation could be handled by subcontracting, with regulated banks remaining the lynchpins. Basically securitization seems like a system for banks to evade capital requirements (and provide fees for intermediaries). To some extent if the bank is risk-sharing and borrowing long-term the existing capital requirements are probably excessive. So perhaps the way to address securitization is to have everything carried on the books of the banks but to allow them some reductions in capital requirements when they use long-term loans and risk-sharing contracts.
this is why i don't think investor paid ratings would help much. there would still be too many investors who don't know what a dumb loan is and who don't do their own credit analysis.
Excellent analysis. I do tend to agree with Andrew about "burying the lede," but you aren't trying to write in an inverted pyramid, Journalism 101 manner, are you?
Also, I'm beginning to think about an even more fundamental question: Is U.S. residential real estate really the best destination for so much of our (and the world's) investment capital?
What if we developed incentives for investing that capital in productive and life-sustaining ways?
As an aside, reading about crappy origination channels reminds me of a conference I attended maybe 2 or 3 years ago. Anthony Penington-Cross, one of the top academics researching subprime, was presenting a paper that found that mortgages from third-party originators had much worse performance than similar mortgages originated in house. None of the other economists in the room believed his results. The attitude was, 'sure there might be some difference, but it can't possibly be as big as what you found, or no one would buy the stuff.' Draw your own conclusions.
The solution is for investors to refuse to get within 20 feet of a mortgage-backed security that is backed by dumb loans.
Wasn't that the whole sell of the securitization though? The line that the risk was spread out so much that it didn't matter if some of the loans defaulted, the overall result would be to spread the risk thinly? Sort of funny (in not a haha funny way) this seems now.
Unfortunately what happens when businesses make decisions based on what the street expects and on the stock price (if the CEO of a company you invest in EVER says ANYTHING about stock price, sell immediately) and so you are sucked in. If you didn't do these loans then your numbers would have tanked and you'd be punished. Not everyone has the capital or wherewithall of Warren Buffet. So this all just fed on itself. The first error is to think that guidelines had anything to do with actually getting paid back. They had everything to do with getting loans done and fees collected to keep the numbers where they were expected to be.
And frankly investors deserve what they get because they, in the end, drove this with greed seeking unrealistic sustained returns. That is why any bailout drives me crazy. Everyone is the chain needs to come to Jesus from the banks, to the IBs, to the guy with a 401k who believed that he was entitled to 15% returns on everything forever.
Oh boy, I think I had some of that coffee too....
Fair Economist - the argument for securitization is that different parties have different risk trade-offs, and creating a complex security allows you to tailor the risk to the investors. Can you stand a lot of optionality - then buy the last tranche. Need something short-term and safe - then buy the first tranche. Know how to handle market risk, but not credit risk - do a swap with a GSE. By bringing in multiple participants with diverse risk tolerances you expand the pool of people willing to provide capital to the mortgage sector. That's the theory.
I have my doubts in practice. Yes, I think it works that way. But would it be better and more transparent, as someone here (bacon dreamz maybe?) to just have derivatives and not move paper around? Can't handle credit risk - buy mortgage insurance. Can't handle optionality - buy swaptions. The derivatives may actually be a more transparent and accountable mechanism than the rocket science securitization.
So, what exactly is the benefit of complicated securitization deals over just having banks issue mortgages with money borrowed via long-term instruments?
That's another problem I have with the rest of the Davidson piece, which I didn't get to today. The argument is made to eliminate senior/sub structures and rely on guarantees. Really, that is a way of saying that there is no particular benefit to securitization of loans, since they can't stand on themselves as investments, there always has to be some kind of external credit enhancement. At that point, it is hard to distinguish between securitization and on-balance-sheet debt financing of the loans you originate.
Remember that the GSE model, which started the whole securitization of mortgage loans, had something in it besides a credit guarantee: it "homogenized" the loans. Maybe people don't realize how wild west things were before Fannie and Freddie started standardizing. For instance, there didn't used to be uniform notes and mortgages: if you got a loan at Podunk, you signed a five page note that had God knows what kind of terms in the fine print. Decades ago the GSEs got on this problem and published uniform instruments that everyone has to use if they want the GSEs to buy the loans. That was an enormous benefit to consumers.
And so it is with GSE underwriting guidelines: they define what white bread is. Now, there is and has always been an alternative for someone who doesn't want or can't get a white bread loan, but you paid for it and investors in GSE securities didn't have to worry about the things being hidden in a big pool. Institutional investors didn't have to do drill-down loan-level due dilly on guidelines, because GSE loans were "fungible" by definition.
That kind of securitization makes economic sense, and it makes sense that such loans should be allowed off the originator's balance sheet. Certainly the GSEs have the risk--and it has to show on their balance sheet--but for a long time that risk was pretty transparent and relatively easy to analyze.
What we're seeing here is the problem with pooling dumb loans: they perform in aggregate like they do individually, badly. All we managed to do by securitizing them was to concentrate the risk.
And while I'm on this soapbox...you see, writing mortages is a good business. Done correctly they are relatively risk-free, provide a steady income stream that can be counted on, have a very clearly calculable NPV, and as a bonus the servicing end provides jobs and is also a nice business. Boring but nice.
So my prediction is that these boring and nice businesses are going to revert to being local. If I were at a local/regional bank now, I'd be positioning myself to become the lender of choice in that market right now because I could actually KNOW the customers at the branch level and write a whole lot of nice, boring loans that I could either sell to someone that understands the concept of returns, or keep on the books. Just like the old days...sort of like breaking out those old ties when they become in style again.
Another good explanation and analysis but now we are in the soup and there does not appear to be an easy way out. I wrote a paper of the Panic of 1893 nearly fifty years ago when I was in grad. school. I need to look at that period again because it might give some hints as to to how debilitating a long period of deflation might be. It is starting to bite even in the fly over states where I live. I think World War 1 might have evolved from that period which is not encouraging.
By bringing in multiple participants with diverse risk tolerances you expand the pool of people willing to provide capital to the mortgage sector. That's the theory.
And it's actually a great theory as long as the underlying loan behavioral models are correct. But what the models didn't take account of (and probably couldn't and never will) is what happens as this scales up with very large numbers affecting external forces. I think Tanta's post is great at noticing those external forces - the external demand for the securities and the associated fees served to lessen the standards which, in turn, made the underlying models break down.
Which is why I said that, in the end, the only behavior that supports the underlying model lies in smaller scale, or more local expertise when it comes to the mortgage market because, in the end houses, like politics, are local.
Some bankers saw the risks and said no thanks:
'Old-fashioned' banks manage to avoid credit crisis - USATODAY.com
Who behind the curtain knew what they were doing?
Who wrote those underwriting guidelines of no doc loans, with 2/28 arm resets, with no money down, that a bunch of originators were following?
I want to know.
Who behind the curtain knew what they were doing?
You see, that's what I'm saying...NO ONE. If you had to write a financial model describing house prices before this happened what everyone said was true: house prices do not decrease on a large-scale level. Sure, there may be local market exceptions but if you bundle everything together...there you go. You can't lose. Except, of course, unless you think about treating all markets the same.
So think about it. If housing, historically, had always risen 3 to 4%, it makes sense to require less down because even if the loan defaults in 3 years you still make money. It's the way you'd model it.
And I'm sure that someone like me at one of those meetings asked "what happens though if the rate of appreciation slows down for an extended period or even goes negative?" and then got stared at, and blustered against, by the Finance guy because, you know, thy that hasen't happened since the Depression and THAT'S not going to happen again...
Who wrote those underwriting guidelines of no doc loans, with 2/28 arm resets, with no money down, that a bunch of originators were following?
Merrill. Goldman. Lehman. Bear Stearns. Nomura.
Countrywide. Wells Fargo. WaMu. BoA. National City.
Anybody who ever purchased whole loans in order to back a security wrote guidelines. The big guys and the smaller guys.
Why is Luminent or American Home a total failure? They wrote dumb loan guidelines. Why is Thornburg still in the game? They wrote smart loan guidelines.
You want to know who, individually, wrote these documents that are used on the secondary market? I personally wrote more than a few of them. I gave up that line of work when they wanted me to put the 100% no doc shit in there, but yes, I was once paid ludicrously small bucks to write credit policy. And I was hired by the buyers of loans, not the sellers of loans.
Your question is kind of like asking, who wrote the Fannie Mae guidelines? Fannie Mae did. Who wrote the Wall Street guidelines? Wall Street did.
Glad you mentioned how well the old vanilla model worked.
It seems to me that people went from buying anything to buying nothing.
I don't see why we need to worry too much about a repeat of the same thing, since financial markets can quickly go from being excessively lenient to brutally strict. I suppose they could repeat the same scenario, but each bubble needs a new story. Pets.com isn't coming back. There was no need to put in a lot of regulation to prevent the tech bubble from reemerging.
The fact that at this moment there is no active secondary market for true AAA auction rate notes issued by closed end funds at libor + 125bp is shocking to me.
The only reason that we haven't totally melted down is that we have the residue of archaic regulation left over from the GD. We also have the dinosaurs like the GSE's still open for business.
I'll be interested to see your further analysis, Tanta, because I'm inclined to agree with the contention that senior/sub for mortgages doesn't add anything other than what you could get by having the bank issue high-grade long-term debt for the AAA tranches and sit on the rest. Even for the bank, valuing the sub debt is extraordinarily difficult. For anybody else it's virtually impossible. Having other entities "make a market" for sub debt is like making critical business decisions by dart-throwing monkeys.
Fair Economist:
I don't see any problem with relatively simple securitization. I agree that there is no benefit from overly complicated securitization. I don't think we will be seeing cdo^2 again anytime soon.
The whole edifice was based on misrepresentation of risk at every stage of the process. The final one being the rating agencies. I don't see why most of the value (assuming there is any) couldn't be extracted using basic structures.
I'm inclined to agree with the contention that senior/sub for mortgages doesn't add anything other than what you could get by having the bank issue high-grade long-term debt for the AAA tranches and sit on the rest.
Well, sure, but mort_fin's point is that then all you get is the option of buying some pro-rata share of a single-class pass-through with a long maturity and average returns and not much if any credit risk and lovely negative convexity and . . . all the other pleasures people used to get with GSE and Ginnie Mae MBS until they decided they could structure more choice.
The whole argument was that there was this market need for the mezzanine stripped floater piece. Mortgages needed to fill the need for hedges, not just for long positions.
If investors are happy going long on good old single-class pass-throughs, well, fine. Those, we know how to do.
Do we know how to manage the debt issues to keep the asset/liability and duration gaps straight? I dunno, but a lot of folks spent the last several years arguing that that's not really possible. Of course, that's because you're describing the GSE business model and lots of folks just had to argue that it wouldn't work.
Isn't there some space for regulation here?
Make the reserve requirements for Megabank warrantying 100% CLTV, no doc loans so high that it chooses not to get involved? Even when greed takes over, which it will.
In reading this, I am reminded of how Lehman and ALS were attributed with "doing subprime the right way" and of being "best in class" - and how those statements were used to justify the worthiness of these loans. With the benefit of hindsight, one certainly sees what foolishness that was. However, was there ever a way to do subprime "right" and all that happened is, as you suggest, irresponsibly excess supply simply chased the last marginal borrower too far?
I gave up that line of work when they wanted me to put the 100% no doc shit in there
Paint me naive, but I never did understand no doc loans. If you can't produce any docs aren't you just admiting that, on some level, you are lying about something? I mean, if you are hiding taxable income, you are lying to the government. If you're income is uneven, that will come out in the docs or there are ways to get around that, but NO docs?
Is there some place where no docs make sense?
Is there some place where no docs make sense?
Is there some place in residential 1-4 mortgage lending? No. Has never been, is not, will never be.
Does it make sense in the high-rate private-lender hard money loan market? Sure, why not? If you're just lending on the value of the collateral, why screw with the rest of it?
However, was there ever a way to do subprime "right" and all that happened is, as you suggest, irresponsibly excess supply simply chased the last marginal borrower too far?
Sure there was. There still is.
There is just no way to do subprime lending and offer those borrowers the same loan terms as borrowers who can demonstrate willingness and ability to repay. I am not talking about rate here. I'm talking LTV and collateral quality. At some level, subprime lending is always based on the quality of the collateral and the ability to foreclose at break-even. That means low LTVs and strict appraisal review. Anything else has always been nuts.
I gave up that line of work when they wanted me to put the 100% no doc shit in there, but yes, I was once paid ludicrously small bucks to write credit policy.
did you send them a farewell excel card that had a Pig pop up when opened?
OT - Winner/Loser Department.
People back looking for elusive yield since 2-5 year treasuries paying 2%-2.8%.
Adding Risk (Carefully) to Your Bond Portfolio - NY Times
As credit issues get sorted out, people will start buying yield.
I would be OK with life if securitization was put to rest indefinitely. The outcome is lowering the barrier to entry and driving up housing prices. There's no benefit.
high-rate private-lender hard money loan market
Oh yes, that i get...but in a residential mortgage? OK, so using the rarified air of the executive suite or academia, I can understand the concept that people wouldn't put themselves in a position to lose their home, and the reliance on the FICO to tell you that they had a credit history and so, therefore, understood the concept of credit being paid back. But I can't for the life of me understand why income wouldn't be verified on some level.
And you know, at some level I'm still amazed that people would do this to themselves. Come on, you had to know that you couldn't afford it! Everyone is quick to blame the loan guys, but if you take home 5k a month and you're committing to a 3.5k mortgage payment that is RIGHT THERE in front of you...I don't think that anyone sane thought that people would do that. Not defending here, but wtf?
Securitization makes a lot of sense, for the reason already stated - it brings investors with a wide range of risk tolerances to the market - but the investors did not do due diligence, did not require much larger monthly cash risk premiums when downpayments were reduced, and did not conduct strict sample audits on the loan quality.
What can you do when investors throw their money away? Nothing. Just make sure that they eat 100% of the losses so that they wise up next time.
A serious problem with the existing system of providing mortgage finance is the continuing provision of government guarantees. Between the FHLB and FHA and FNMA and Freddie Mac, virtually all the new money used to finance mortgages is now effectively guaranteed by taxpayers.
As long as a large segment of mortgage finance risk is guaranteed by the government, the only place for private risk-taking is at the margins. Why did all the private money growth occur in the riskiest marget segments? Because if the govt is taking on risk at lower cost than the market, you can only compete by taking on even more risk without charging fully for it. So that's what happened. Surprise!
In a market with securitization and very little govt risk-taking, investors and borrowers would sort out what risks are worth taking, and for how much extra charge. Entities like GSEs could focus on providing "white-bread" underwriting guidelines as aggregators and servicers for private investors, without the guarantee.
and the reliance on the FICO to tell you that they had a credit history
See, that's the problem.
If your credit record is not made up of past no-doc no-down loans you got and never defaulted on, it doesn't tell me how you will handle a no-doc no-down loan.
We had borrowers whose FICOs showed they could handle credit that was granted to them responsibly, and we used that as evidence that they could handle any dumbass set of loan terms.
It's a lot like assuming that anyone who can get an A in high school can get an A in college.
People did it because they were either leveraging a speculative investment that looked risk-free (housing always goes up!) or they really didn't understand how much the payment was.
My all-time fave no-doc loan was the one I had to buy back from the investor (it was just my unfortunate job to have to take that kind of call). The investor picked up the loan in QC, and found that the idiot on our end had left the copies of the tax returns and a couple other docs in the back of the loan file.
The deal started full doc, it became clear the borrower didn't qualify, so the soon-to-be-fired LO resubmitted as a "no doc."
Borrower's employment per tax returns: chartered fishing boat captain. Borrower's source of funds for down payment on home: notarized agreement for sale of a . . . fishing boat.
The underwriter who dealt with it the first time did catch on to the difficulty of selling your only means of making money in order to cover the down payment, and so denied the damned thing. That's why it came back as no-doc and in those days (this was back in the 90s) this no-doc crap was so new we didn't have mechanisms in place to catch these resubmissions. (Note: everybody does now. Every loan origination software out there can flag these things based on addresses or SSNs, etc. So brokers just send it to a different lender now.)
Anyway, the whole point of that deal was that the funds for down payment were coming from illegal activity (that boat was running more than fishing expeditions). In order to launder the money, this fake bill of sale on the boat was submitted.
The system worked originally: that loan was denied. It failed when the no-doc option allowed it to go through the second time.
I will just generally observe that there are plenty of drug dealers with excellent FICOs. Users might hose up their credit records, but the dealers don't necessarily do so if they don't sample their own wares that much.
I still think the underlying problem is devoting too much money to houses. There's a finite amount of investment capital. We desperately need to sink more of this money into transportation infrastructure, agricultural, any number of things we will need to survive as a civilization for another 50 years.
The solution is for investors to refuse to get within 20 feet of a mortgage-backed security that is backed by dumb loans.
The solution is for investors to refuse to get within 20 feet of an american borrower. Deadbeats that will walk away from anything when it suites them. That includes walking away from a house, a spouse, parents, etc. After all if its not right for me how can it be right for you. Ah, yes, I remember my hippie, dippie doo group meetings in the 1970s on open marriage, divorce, and in general if it feels good do it and screw the consequences. Now it feels good to walk. The vengeance of the "Woodstock" generation writ large.
"pink witch"
Ha! The mtg pig appears!
So is Tanta the "good witch" or Dorothy?
George W is obviously cast as Scarecrow.
Lion - Ben Bernanke? He's kinda furry.
Tin Man - Paulson? He's pretty stiff.
Munchkins - CR community? Or are we the Flying Monkeys?
And is CR himself the Wizard?
Oooooooo....
ok...I've heard the boat story...
tell us another one !
tanta
maybe it was a sale leaseback of the boat
you have to consider ALL the possibilities, you know.
maybe it was a sale leaseback of the boat
you have to consider ALL the possibilities, you know.
That was exactly what the dude said when the first underwriter said no deal.
So then the underwriter went through and updated the cash flow estimates on the business with the lease payment in there. Ooops, ratios didn't work any more.
We weren't always this stupid.
lol...I love those stories Tanta. I have tons of bad business decision stories. Once you sit at certain levels, you become further amazed that anyone actually makes any money. Having worked on a team that did an on-line banking implementation (one of the first), the stuff i saw makes you understand how all this bad crap is just business as usual at banks. I'd be shocked if at Citi, the software even now could aggregate all a customer's data in one place at one time.
When i retire, i'd love to teach at a b-school and make some of the stupid things i've seen part of the canon. The events, and the people, deserve to be part of history from the inside perspective
Who was that fisherman?
His name wasn't Peter Griffin was it?
I remember my hippie, dippie doo group meetings in the 1970s on open marriage
The hell you do.
"Group meetings"? (What other kinds of meetings are there?) "hippie, dippie doo"? Where'd you pick up vocabulary like that? Old Hee Haw reruns?
Woodstock was 40 years ago, dood. You gotta update your strawman. Maybe you could listen to Mitt Romney do gangsta for some new phrases. Wassssssuuuup? Who let the dogs out? My maaan!
It may take a few minutes for my eyes to roll back into the front of my head.
In many ways the root of the problem is a reward structure that favours the short over the long term. Write a loan that goes bust a couple of years later-no problemo, you already cahsed your bonus check. Do a merger that 2 years later turns out to be dumber than dumb-same story. Cur research to juice quarterlies at the expense of future product streams-again no problemo.
I would like to see at least a good portion of bonuses/options/executive compenstaion beyond reasonably modest base salaries placed in escrow until the actual results of the loan/deal/whatever are known. For mortgage loans that would be tied to repayment.
The whole argument was that there was this market need for the mezzanine stripped floater piece. Mortgages needed to fill the need for hedges, not just for long positions.
Ah, but the problem is that nobody can give a reality-based value for the mezzanine and equity pieces. A few percent change in the value of the pool - a sneeze in the credit market, really - and the mezzanine goes from a nice investment to a deathtrap.
We've got a really intense demonstration right now that when you chop up these pools into little bits you can get in a situation where nobody can possibly know what any individual bit is worth. We have created an enormous amount of uncertainty. And precisely because the valuations are so unclear I think they make really lousy hedges regardless of the fantasy models the quant made. It's the old idea - you can hedge risk but not uncertainty.
Now you could issue the AAA tranches, no problem - basically a loan to the bank with collateral. Securitization in that sense is practical. Likewise the bank could issue a complex security for hedging if there's a big demand. But I don't think the junior tranches are a good financial instrument and especially not when in paper-thin slices.
Tanta...
I think you hit the nail on the head, kinda sorta.
Most of the ultimate investors in securitization tranches are institutions or the equivalent thereof who simply WILL NOT (not that they could) do the credit analysis themselves; they have to rely on the credit, ummmm, "rating."
At the end of the day, the credit rating depends very largely on what the model assumes of owner behavior.
Yes, there are exogenous variables in the ratings models, such as the probability of refinance given what yield curve exists in the future and so on (one good reason that attempting any real precision in this is ridiculous)...
But as you kinda sort say, we have "good" information about how homeowners historically have behaved in a wide variety of scenarios WITH NORMAL LOANS, like 80/20 30 yr fixed full doc.
I don't think it's enough to worry about fraud in the apps at all, so assignee liability is not so important as realizing that while its true the investing public has a short memory, who in their right mind would ever buy a tranche again in any trust made up of weird s--t? I mean its obvious you cant trust the ratings agencies even if they WEREN'T conflicted.
To that end, wouldn't you agree that any stupid plans to bail out the (bonds, homeowners, banks, you name it) would have an aversion consequence, as the moral hazard such could represent could corrupt otherwise normal homeowner behavior?
A few years ago it seemed like reduced doc loans were the future of the mortgage business. No matter what rationilization you use to justify their validity they are tough to get right. It's too easy to get taken to the cleaners.
Alan Blinder wants to revive the HOLC from 1933. He believes it could actually turn a small profit for the Treasury as its earlier version did even with a 9.6 % default rate.
ECONOMIC VIEW; From the New Deal, a Way Out of a Mess - NY Times
Very interesting post and some thoughtful comment threads.
I am in favor of a more localized lender for all the obvious reasons. But you cannot discount the IB's wanting to get their greedy fingers in the pie.
I think part of the issue yesterday and today is one of capacity...there are simply too many folks at every level in the game. That's probably what drove development of some of the insane guidelines but more importantly, bastardization of the existing guidelines (that worked) from way back.
If there is to be more regulation, it ought to center on how to limit the players at every level. State licensed brokers are kind of a joke as are the New Century's of the world. But if a broker and a company like NC wanted to get in the game, sponsorship from a GSE-like authority with a rather high barrier to entry might make some sense.
GSE's work because the guidelines make sense and everyone selling to them knows they had better not be loosey-goosey with the deal or they will be out of biz in a heartbeat. If a broker or a New Century was under similar scrutiny, maybe today's issues wouldn't be so pronounced.
To quantify what I just said...
You can probably always have AAA ratings on the top 50% (maybe) no matter what (within reason) weird stuff is in there as long as theres no rampant fraud...
(Just hypothetical numbers, but you know what I mean)
But the ratings would have to fall off a cliff below that, so instead of 92% AAA/AA you'd have more like 15% BBB and 35% unratable.
In a shift like that, assuming AAA needs 35bps on average and BBB needs 450bps spreads, and 35% is either unsellable or needs 1500bps spreads, then the interest cost on the loans has to be way higher to support that, probably 450bps or 550bps higher (that would be an eyeball estimate of the extra spread per loan [on average] that you'd have to charge on anything other than plain vanilla loan securitizations).
And that would eliminate the teaser stuff because the spreads would be forced higher to compensate for the risk.
Which is as it should have been.
What were they thinking?
A mortgage insurer claims representative was berated by one of the Wall Street suits whose sleaze ball aggegators had been inhaling as many of these crazy loans as they could find, complaining about the huge recission rates on claims from his pools.
The representative reminded the suit that mortgage insurance coverage was contingent on his organization's reps and warranties and that the coverage had a contractual exclusion for fraudulent originations in which the originator contributed to the fraud.
The suit then had the chutzpah to argue that the MI was selling fraudulent insurance, and that 'everyone knew' that a 'certain amount of' fraud was one of the unavoidable costs of doing this business. 'So either make good on these claims or we cut you off from future MI RFQ's and sue your behinds off for bad faith claims practices.'
MI claim rep: "Sue away."
The suit made good on his threat to cut off subsequent RFQ's but nothing more was heard about litigation over the denied claims. This was two years ago. The reps' company's only regret now is that this conversation did not take place three or more years ago. Quite likely, many more of these crazy loans were originated fraudulently, but without the investigation bringing to light proof sufficient to prevail in "bad faith" litigation, the claims get paid.
When large numbers of these fragile transactions are concentrated in a community or neighborhood, all of the owners of nearby properties, regardless of the quality of their mortgage debt or their outright ownership, are penalized by the market disruption resulting from very high REO and property abandonment rates. I suppose this is the point of Tanta/CR's slogan "We're all subprime now". Holders of MBS backed by "prime" loans will be slammed as bad as the ABS, if their collateral is secured by property in severely declining housing markets. The prevalence rate of subprime and crazy pseudo-prime (Alt A) lending is unprecedented, so the losses are also likely to break all records (at least all records since the Great Depression).
What were they thinking?
Do you really have to ask that Fred?
Although, if you go back to before the numbers got to a certain point and look at the historic percentage of foreclosures, ALL the numbers would make sense, and the point spread i don't think has to be as severe as you think, especially if you look at pre-bubble data when most of these models were worked out.
Aheadofthecurve writes:
"Alan Blinder wants to revive the HOLC from 1933. He believes it could actually turn a small profit for the Treasury as its earlier version did even with a 9.6 % default rate."
The embedded default rate in the subprime and Alt A books, especially in declining markets, will easily run north of 20%. Even in a flat market a lot of this crap would be expected to run default rates well over 10%.
During the 1980's oil crash, loss rates exceeded 30% overall in the most affected markets for LTV>80 debt, and ran 40 to 50 percent for the debt that had been originated with riskier terms (investor, neg am ARM, declining market). This time, it is not external economic shock that is trashing housing prices, but the implosion of a reckless housing finance induced bubble.
These bailout schemes are pure madness, just like the HF madness that got us here in the first place.
o thats the point spread you'd need for anything much weirder than 80/20 full doc fixed rate.
how can anyone know what HONEST homeowners in a 2/28 will do in various exogenous scenarios?
that was my point, we have no data on which to model behavior.
CSO structures should be safe because we presume that with 80 years of default and severity data and putatively responsible managers we can be pretty sure of hazard functions.
but we have no data on what people who own new s--t will do, and no data means wide variance in the models, and that means any investment grade rating is going to have to have a big spread to cover the tails of unknown, but real, fatness.
I sell protection for private clients on CSO tranches because I can run my own models and do a reality check and I can pick the names that go in there and the WARF and so on. I sell on risk-adjusted return with fixed (specified digital) severity.
But I'm a fiduciary who's not in awe of and not buffaloed by ratings.
I have no clue how to handle alt-A pools or 2/28 pools and so on, nor I suspect does anyone.
So you'd have to pay me HUGE to take the model risk.
Do you agree?
"Do you agree?"
I agree that when all this crap was being generated, the model default rates had to have been based on extrapolations from tamer stuff - - -the crazy terms were unprecedented.
So, yes, the "model error" will be humongous, little better than a WAG, no matter how fancy the MC simulations and statistical analysis of the historical behaviors that were embedded in the pricing and rating models (and, from what we have seen of the rating models, they were often anything but fancy - - - incredibly crude!).
However, we now have million of such loans with developing experience that will be great for future modeling - - - because they are aging through increasingly stressed markets. The only fly in the ointment from the perspective of behavioral modeling integrity is that the implosion of originators and financial and operational stress on the servicers is likely to induce a system-wide case of institutional Alzheimers, and the behavioral data will not be properly preserved and archived. If they can't even keep track of the paper chain of ownership, what hope is there of capturing and preserving the underwriting data that supported the funding decision? Too bad!
"If you do not know what a dumb loan is, you might want to consider investing in a different kind of instrument."
I'd go much further. If you do not know what a dumb loan is, you should immediately choose another line of work.
Blue Ridge- Sorry I misspoke on the HOLC. Default rates were 20%; loss rates were 9.6%. I don't know what default rates would be here. I would think it would depend greatly on what loans are taken in to the program.
If you're at a poker table and you don't know who the chump is, it's you.
Apropos.
If HOLD defaults were 20% and losses were 10% and losses are almost only due to defaults then help me here.
In the 1930s the highest LTV was typically 50%.
If 50% LTV loans produced 10% losses then one of two things must be true:
(1) There was a bubble then even worse than this one;
(2) A 21st century HOLC will be killed.
Since all reports say we're in the worst bubble in 100 years (well, a lot of them say that) I'm not inclined to believe (1).
So if a stupid HOLC plan ever shows up and it involves short refis at 100% of current market value and we're not through the deflation yet...
How stupid is that if the LTV on those loans is by definition 100% and the LTV in the 1930s was half that and there were losses?
Wouldnt buying bonds to back that be insane?
So the gubmint is now going to sell even worse securitization stuff (bonds) than the street ever did, and to retail yet?
Tanta,
Nice -- worthy of inclusion in the UberNerd files.
Quite a while back you made a statement that's stuck with me. Basically, you stated that non-GSE mortgage securitization is likely not (and probably never should have been) economical. I believe that's the absolute truth.
As you've noted, securitization works for the GSEs because of their strong guidelines and the uniform nature of the underlying mortgages. No one can compete with the GSEs on their own turf, so what's left? Subprime, ALT-A, Jumbo, etc. -- stuff that defies uniformity, engenders much greater risks, and will necessarily involve much greater processing costs. It's highly unlikely that securitization can account for all those additional costs (plus those imposed by the securitization itself) and offer competitive rates.
I'd love to see you (and MaxedOutMama & bacondreamz & others) run the numbers again, but I just don't non-GSE securitization coming back... at least, not until this generation is dead and buried.
p.s.: Unfortunately, the IBs appear to have spoiled the party for everyone. The blowup and collapse of the mortgage bubble will ultimately take down the GSEs themselves; those "agencies" that would've otherwise been fine are now going to be taken down with everything else.
fred-I suppose the discrepancy between default and loss rates could be due to the defaults being p[redominantly at the lower end. Don't know for sure.
I'm not sure there would be specific bonds for HOLC. Rather the money would come from normal treasury borrowings. That would allow the borrowers to pay a lower rate than otherwise while still allowing government to make a profit.
I think there are 2 classes of borrowers out there, amongst the non-speculators who actually want to stay in the homes:
(1) those who can't pay at any conceivable interest rate.
(2) those who can pay at 4-5 % but not at 8-9%.
(1) are cooked regardless. I'm guessing Blinder wants to target the plan at (2).
I don't ask many questions here - but I asked one yesterday. It wasn't answered. Perhaps it got lost in a very long thread.
I follow a blog called the Dream House Diaries. From the NYT. One of the married bloggers writes a financial column for the NYT.
They bought a lot in Florida a while back for cash. Started building last year solely on credit. At the time - BOA gave them what looks simply like a HELOC with a $1.3 million LOC. At 7.X% interest (ARM adjustable starting in 2012). There is no fixed amount on the mortgage as it is recorded in the public records.
What happens when construction is finished in a few months? Does the HELOC for construction purposes convert into a loan at a fixed amount? Or are they subject to BOA rethinking the HELOC? I'm a lawyer - and this document doesn't look to me like the bank has committed to any permanent financing. The bloggers seem to think it has (although there is nothing to that effect in the public records - perhaps the bank's "promises" don't have to be recorded). Are there construction loans like this? What usually happens? And what is happening these days with these kinds of construction loans? Finally - this house will probably be worth less than it cost to build by the time it's finished (it's in Manatee County FL). What happens then? Any and all responses - even partial - would be appreciated. I really don't know beans about mortgages. Roby
robyn
i had clients who did deals similar to what you described with the private client group of bofa.
realize im in ny not fla.
nevertheless, the paper in my case did not contain any legal obligation, contingent or otherwise, on the part of the bank to convert to permanent financing upon completion. in fact, most private banking loans like that demanded a takeout within 2 years no matter what.
of course, ymmv.
BTW - regarding mortgage backed securities. I don't know anything about taking out mortgages because I haven't had one in about 30 years (just kept rolling sale proceeds into new places - once every decade or so). But I do know about fixed income investing (our portfolio is 95% fixed income and we have been living on it exclusively for over 20 years - until my husband started to get SS last year). For almost 30 years - I have read about investing in mortgage backed securities. And other esoteric stuff (like auction munis). Fell asleep many a night reading through several editions of Fabozzi. And you know something - I never "got it". Never understood a single one of those convoluted securities. So I didn't buy any.
I did understand yield curves and interest rate cycles - so I simply tried to buy mostly plain vanilla stuff when I thought interest rates were high - and avoided the same stuff when I thought interest rates were low. This is something an institutional investor or a hedge fund cannot do. Because those people are always "on stage" - and their performance is judged quarter by quarter - even month by month. If interest rates are low - they have to do something - anything - to juice up returns. Which has meant - in recent years - buying stuff that they didn't understand - frequently with lots of leverage.
Anyway - I think all of you can try to understand these securities - but - bottom line - I don't think anyone can. Except in the most basic terms - and almost all of these securities have failed "stress tests" when subjected to real life less-than-optimal market conditions (we had it on a smaller scale with LTCM - and we're seeing it now on a much larger scale - Greenspan's fault - if you don't let a smaller stupid thing fail - you wind up with a bigger stupid thing failing).
I will give you a simple example. I own "10/30 yield curve inversion CDs" (a small side bet on the yield curve - heads I do ok - 8% - tails I get less - but my principal is insured by FDIC). I know why I bought them - but why on earth was HSBC selling them to retail investors? I am sure that I and HSBC are not the only parties involved. Who is holding the (probably leveraged) third and fourth legs of this bet on this transaction - in what form - and why? Roby
Hi Fred - Well this is not my mileage - it is the mileage of the bloggers. Who also dealt with the private client group at BOA.
They have been so far off base with so many aspects of the construction that I wouldn't be surprised if they were off base here too. FWIW - this is their third house (each owns a separate house up north). And I seriously doubt they can afford this third one in the absence of "silly money" financing. Which I think is gone in Florida now best I can tell.
Usually - anything a bank promises is not only in writing - but recorded in the public records of Florida - and there is nothing like that here. Will be interesting to see what happens - especially since one of the bloggers is a financial writer for the NYT. Roby
robyn
hsbc is notorious for stuffing this kind of thing on the retail channel. its VERY VERY profitable.
two points:
(a) i have clients with nine-figure portfolios who get pitched this stuff by private client groups and as a quant when i reverse engineer i find that the built in profit is unbelievable.
in your case, you got a cd that instead of paying X% pays either 8% or something else depending on the slope of the far end.
i GUARANTEE you that the cost to hsbc of providing this...remember they are not warehousing the risk, their desk hedged it away...is LESS than it would cost to provide a plain vanilla cd.
in other words, if the going rate for their cds was 5.0%, i PROMISE you that the internal cost to them of providing this thing to you was 4.5% or lower.
(b) on the other hand, as a retail player, it would be very hard for you to take such an exposure on your own, let alone at wholesale rates. in this case you could probably do something similar by rolling bond futures, but you'd have yourself a tax straddle there and other messiness and its not for amateurs. so in that sense, its not fair to say you were hosed, because there's no real likelihood that you would ever do this yourself, as you admit (no criticism intended).
structured products in the retail channel are a dirty little (not so) secret, hugely profitable because the retail channel lacks the quant skills to know the built in profit and even if they did lacks the ability to do it on their own
(for one thing, by cftc regs you need $5M investible to transact otc derivatives for hedging and $10M to do so for speculating).
does this make sense?
as to "understanding" the weird stuff, some of us DO understand it, but what we understand (if we're smart and honest too) is that the number of impossible to estimate cross currents of things that can go wrong make it an uncertainty, not just a risk...
uncertainties are by definition non quantifiable and to be avoided.
your point as to on stage is exactly right, as ive written; these people are graded on relative performance and can't try for opportunity.
robyn --
as an example in cre...
i had a client who got a mortgage approval of $10M to buy a property and didnt want to close for an extended periof of months while he did a due dilly.
the lender quoted him $200,000 for the rate lock.
i was able to go on the cboe and replicate a rate lock for him for $80,000. could also be done otc with a desk if you know what you're doing.
i PROMISE you unless they're brain dead the traders at that bank would have done the same thing and kept the $120,000 as profit.
same thing with your structured cd.
"i have clients with nine-figure portfolios who get pitched this stuff by private client groups"
Guess whos going to have to sit on a jury and listen to this stuff get litigated. It would give me a bad case of narcolepsy.
im not my clients. you'd be amazed what stuff is pitched. the richer they are the more weird stuff gets shown to them.
"John Stark writes:
I still think the underlying problem is devoting too much money to houses. There's a finite amount of investment capital. We desperately need to sink more of this money into transportation infrastructure, agricultural, any number of things we will need to survive as a civilization for another 50 years."
I think this nails it. Things are rapidly getting more expensive, like oil and food. We will have to live in shacks to afford to eat and get to work.
"You want to know who, individually, wrote these documents that are used on the secondary market? I personally wrote more than a few of them. I gave up that line of work when they wanted me to put the 100% no doc shit in there, but yes, I was once paid ludicrously small bucks to write credit policy. And I was hired by the buyers of loans, not the sellers of loans."
If the persons who wrote the egregious credit policy are not held personally accountable, then there is no moral hazard at all in this business. They did what they were told was good for business and moved on to greener pastures.
Under this side of my coat I have some nice looking watches, says Pulsar right there.
Fred - Actually these structured retail products have been much to the advantage of me and my husband - and my late in-laws - and my father (I manage money for all of us). Callable CDs (basically used to finance mortgages best I can tell) have been the best - yielding 100-150bp or more over treasuries. Yes there is interest rate risk. But they are CDs - and puttable on death (which I did with some when my in-laws died).
I know that the brokerage firms get about 3% commissions. Don't know about the outfits that issue them. But whatever they make - it doesn't bother me if I am getting "rich" yields on FDIC insured products which are relatively easy to understand. Especially since I can buy them on line. Very easy transactions.
I also do a fair amount of muni business. Mostly on line. A lot more work in terms of credit quality analysis.
BTW - brokers have pitched complicated stuff to me too. But I like to KISS. Roby
Anonymous - Don't worry about falling asleep. Most cases involving client/broker issues wind up in arbitration. I was an NASD arbitrator. Gave it up. The pay was lousy and it took me 6 months to get my travel expenses reimbursed (used to be only 2 arbitration venues in Florida - which is a pretty big state geographically - that's still the case to the best of my knowledge). IMO - it's a crummy system. Roby
We've been down this path before...ie assymetric information, principal-agent, adverse selection, moral hazard, etc. This is the old lemons issue. It's all about incentives. Everyone in the daisy chain of fools needs to have some skin in the game otherwise quality will slide (especially when the fed is giving away money). Some sort of assignee liability "lite" should be seriously considered because even though wall st. delegates the decision, if wall st. was on the hook this stuff would not have been originated. On a separate topic, who thinks we are in a liquidity trap? We're not there yet...but stay tuned.
I do not understand why we allow non-GSE-backed securitization at all.
In the recent schemes, investors wanted to participate in the "huge" potential of doing business with consumers--without any of the regulation that normally oversees most consumer financial services businesses. These regulations prevent excessive externalities, fraud, and all the nonsense we suffer from today.
Normally, if you want to participate in that party without any personal liability (or accountability), you buy a stock or an on-balance-sheet bond issued by a major retail financial services firm. In return for limited risk, you give up all day-to-day management control. An obvious and rock-solid principle of corporate law.
But these investors wanted to interfere in the running of the business, by offering products no responsible player originating for its own book of business would ever touch. And they want to control ongoing servicing of that business by virtue of their agreements.
And I'm supposed to believe it's a benefit that securitization attracts such players to the market?
Interesting side note in response to Tanta's statement -- "I will just generally observe that there are plenty of drug dealers with excellent FICOs. Users might hose up their credit records, but the dealers don't necessarily do so if they don't sample their own wares that much."
In my last origination position, I was approached by an Account Executive for a now-defunct-lender who bragged about how committed his company was to funding stated income loans. He then bragged about how one broker had submitted a stated income, stated asset (dumb) loan to him for a 100% purchase and listed the buyer's employment as "drug dealer." Instead of reporting this to law enforcement, the processor at this lender simply called up the guy's arrest record and based on the consistency of the charges levied against this guy, confirmed his "employment" and the loan was funded.
When you had this kind of activity going on for the last several years, how on earth can anybody be surprised at the mess we're in?
Some (more) Facts:
1) All if not most of the instruments could only be purchased by 'Qualified Investors'/sophisticated investors.
2)Securitized & AsstBacked-Debt had been done, historically, with 'safer'assets and better credit quality. The toxic stuff they included in many of the more recent pools was crap.
3) All the institutions Knew they were selling each other stinky-toxic stuff, now they have No Trust for each other nor do they have a clue of the PV or FV of their holdings.
4) Did anyone notice the credit default market growth(simply, why so much risk aversion if you are not making dumb loans?)
5)Banks, HFunds, pensions, & MFunds all wanted to juice their return while appearing to keep the Beta of the portfolio low(er); you look smarter that way.
6) Greed, at every level.
Banks make money when they make a good loan. Banks make money when they make a bad loan. Banks make money when tax payers have to bail-them-out (or get rate cuts). hmmmm...,too bad I can't screw people over, if not, I'd be a Banker.
We need to do what we told Japan some years ago, let the bad banks go under.
...wait til' credit cards really start to bounce.
-cheers