I started looking at the NDE PR this morning and had to go get dressed and take a walk.
I really always had some respect for Michael Perry. But this kind of "overdisclosure" that actually conceals "underdisclosure" in a sea of numbers is pissing me off.
I didn't understand it the first read through. I'll do a REMIC Remix at lunch and see if I can digest it then.
Alas, my Mom never played Mozart for me.
why did Grant Thornton just walk out on a couple of its auditees when the subject of secondary market pricing strategies came up? Are all these Wall Street-inspired conduits really paying a perfect price for this stuff?
seems like some of my comments were cut off. I knew Grant Thornton (of Parmalat fame!) had quit Accredited and Fremont but didn't understand your aside.
I assume it had something to do with either kickbacks or not marking to market properly (or at all). Could you please explain?
lama, I really really want your final feedback on this. I'm willing to take 100% responsibility for not explaining things well. But I also, assuming that my communication skills aren't that awful, want to know whether a person such as you, with above-average financial skills, can go along with me here at a summary level.
I think this is important for two reasons: one, I'm just talking to myself about some issues if a widely-drawn readership of financially sophisticated people have been so unexposed to the structured MBS market. That may be why my personal skepticism about the extent to which a lot of risks have been truly "dispersed" doesn't make sense to other people.
Two, if someone with a CPA and better than average attention span finds it mind-numbing (and it is), then it's just like conversations about neg am ARMs. The fact that someone is taking one isn't saying much about why someone is taking one. Same goes with the securities we put them in. There are some sharpies buying stuff for hedge purposes, some institutional investors who are just buying a rating, and then some other group of investors who, frankly, may be about as up to speed as the Option ARM borrowers. That idea, if it's true, should frighten everyone. It frightens me.
Tanta - I bet you could put this article to a great Dr. Seuss limerick. It also might make it a tad easier to understand for some of us who are closer to Dr. Seuss level than to yours...
Everything has gotten so amazingly complicated, it can be overwhelming. The thought of escaping to the woods is a temptation, except that the developers already have their eye on that parcel and the property taxes are unaffordable anyway...
Actually, it was a mere note because I don't know exactly what GT was worried about. But I fear the issue may have something to do with these "held for sale" pipelines.
An outfit like LEND or FMT may do some retail originating, but in most cases they are putting out a price to a correspondent or a broker. What they pay for a loan, obviously, is a question of what they think they'll get for it, either selling it as a whole loan to some other conduit or securitizing it themselves. In other words, there's a "bid-ask" spread in the pricing at this level.
So when I start to see a bunch of parties disclosing some surprising losses on the HFS pipeline, I do not necessarily conclude that the problem is that the market just happened to fall away on these guys and suddenly they can't get the bid they should have gotten on the resale of the loans. It seems possible to me that they just mispriced it up front, counting on the market to keep rising and bailing them out of it. In my experience this happens when someone is desperately chasing market share: it prices some loan at 105 to the originator just to take that loan away from the competition, who is pricing it (perhaps more realistically) at 104.5. You can spend some time doing that if you can keep securitizing what you buy. If the securitization market falls off, and you have to sell it as a whole loan, you get your ass bitten.
This is also a famous trick of outfits (AHM, perhaps?) who have a retail business to support. You have to keep loans coming in because retail is real, true overhead costs. So you get tempted to "subsidize" retail by mispricing. My impression is that that's what happened to one of the first imploders, MLN, although I haven't followed that story fully.
In any case, it can boil down to your auditor forcing you to take impairments on the HFS that you don't want to take.
I thought this was admirably clear, Tanta, and it is a wonderful effort.
Regarding whole loans vs tranched securities, it is worth contemplating that net flow of income to bondholders on tranched securities cannot be more than the net flow on the same whole loan pool. If securitized loans are selling for more overall than whole loans, it is because somewhere, somehow, someone is getting sold a bill of goods.
The essential for securities markets is full disclosure, and one would strongly suspect that full disclosure is impossible for some of these tranches. No one knows how they will perform long term.
This article on REMIC's will take more than a couple of read throughs for me to understand the layering and pricing of these debt/hedge instruments.
But I have a question. Did some of these exotic loans come about because the mortgage refi market for super prime borrowers dry up after the big refi boom during the summer/fall of 2003? Remember that was the time when ten year treasuies hit rates not seen for 40 years. The reason I am asking is because some investor some where, probably in China or Japan, is loosing a ton of money on my 4.36 15 year fixed with 11.5 to go. Having had 14 percent mortgages in the early eighties and having refied many times as the rates ratcheeted down. We like other boomers are now at a point in our lives where you can't and wont move unless downsizing with a complete cash purchase. The inverted yield curve has forced the creation of all these different MBS just to keep a long line of people employed and still in the game. The only problem is that the yield curve was not supposed to stay inverted this long. Certainly this is screwing the models that were used to create these exotic securities.
Thanks, MOM. I think it's quite possible that someone is being sold a bill of goods. I just want to say one thing for today: given how much piling on has been going on, generally, with borrowers who took dumb loans (hey! they got what they deserved! they shoulda paid attention!), I am hereby ready to pile on any bozo who bought a tranche of this stuff. Sauce for the goose is sauce for the gander: the whining from bond-land is already starting about how nobody told them that this deal has gigantic risks. So if homeowners have to suck it up, then the rentier class has to suck it up.
Sebastian, nobody who fails to stress-test things gets to turn around and excuse himself on the grounds that problems having to do with stresses that had a high probability of happening were "unanticipated" or "unforseen."
Did some of these exotic loans come about because the mortgage refi market for super prime borrowers dry up after the big refi boom during the summer/fall of 2003?
Short answer is yes. Exactly. And you can plot that (exotic product share/refi share/share of loans in the money for a fixed-rate refi) on a graph. If I had one that was publically available I'd post it. I may have something I can scrape up on that.
Anybody buying fixed rate paper has to have some kind of hedge on it, or else, as you point out, you own an investment yielding in the mid 400bps when new ones are in the mid 600 bps range. Hedges, though, are expensive. As far as I'm concerned, Mr. Greenspan came right out a couple of years ago and said to depositories, put those customers in ARMs so that the customer takes the hedge. Jerk.
ozajh, it's not clear to me what the answer to that is in exact percentages. The investor can take it or the servicer can, and it will depend on the servicing agreement of the REMIC. I'd say this: it used to be investor money (offsetting the yield problem with a prepayment) in the "pass-through" days. More often now it is servicer money (offsetting the servicing fee problem with a prepayment) in the structured security days. I would hazard a guess that in most recent issues, the prepayment penalty goes to the servicer.
I think I just saw the one that got away. Looks as though Mr Seiders and the HB lobby is getting some face time with the Agencies:
"The loans Freddie Mac buys under this program would not be limited to refinancing, though refinancing is the initial focus now that millions of people have adjustable-rate mortgages with low teaser rates that will soon spike."
So, looks like they are throwing in a little something for the HBs too. I was puzzled at why the HB stocks have gotten a bounce in the last few days.
The market economy, where no well-heeled, sharply directed influence goes unrewarded. The gravy trains keep running. They just never seem to stop and pick me up.
Nice article, Tanta! Thanks for taking the time to educate this one member of the unwashed masses.
I was taken in by the IndyMac article, however; in particular, how factors in the S&P LEVELS(tm) model are literally driving the changes. Leave aside the fact that a regression model based upon backwards-looking historical data is likely to be stunningly less predictive when faced with novel mortgage vehicles with built-in discontinuities. What took me was their faith-based belief that the risk factors were independent, and that tweaking the weights of some parameters of the model would suffice to control other risk factors:
"Lower documentation requirements on loans are more than offset by other compensating factors, such as lower CLTVs and higher FICOs, with the result that lower documentation loans actually have lower estimated lifetime losses."
Another interpretation is that the model underweights the risk of certain documentation types due to lack of historical precedent, making it cheap to offset the effect of that risk in the model by tightening up on an overweighted factor. I hope these securitizer using this LEVELS(tm) model are pricing in "model risk": the likelihood that the next few years data will lead to a greatly-changed LEVELSPLUS(tm) model that accounts for all the soon-to-be-noticed failures of the current scheme.
Very interesting. As background, M.S. in Aerospace Engineering. I think I understand the jist, but I need to make a diagram and some equations to see how this stuff works practically.
What I never understood is where does the neg-am accrual go? They can't be paying money out to the bondholders that doesn't exist, right? Does that mean somebody has signed up to get neg-am payments, and in the interim, they get the interest paid on the neg-am portion each period? Where does the slush fund live? Does the initial premium paid (the 5 in "105 cents on the dollar") give the funding to keep the cash flow from cutting off in the early months?
Until I read Tanta's lucid description of the mortgage repackaging market, I had always felt that there was a law of conservation of complexity in financial markets.
Very complex markets often develop around widely held simple tradables (like shares for example).
On the other end, you get simple markets for really complex things. Considering something complex like power plants for example, there are very few buyers and sellers so you don't really need a sophisticated marketplace.
The mortgage market is the first example that I have seen of a complex marketplace for a complex commodity.
DCRogers, you are so right. I think it's even worse than that, because I don't trust their "CLTV" numbers. I'll try to post something on that later today.
Tanta wrote: "The claim that premium pricing prevents rate predationthat the investor doesnt gain by having a borrower pay an above-market rate because it pays too much premium for the loanis a mite disingenuous in the context of structured securities. If you are buying whole loans for an investment portfolio, or for inclusion in a simple pass-through, that might make some sense. But if you are buying loans for a security in which someone will benefit from the fast prepayment of that loanand someone else will benefit if it stays on the booksyou may well be paying up for that loan precisely because it has the payment and prepayment characteristics you want, not in spite of them. There is always someone on the other side of a hedge trade. If there were no incentive for conduits to pay 105 for a loan, I can assure you that they wouldnt pay it, or not for long."
The claim is that the rate the borrower pays is irrelevant for the wall street conduit or any other investor, in terms of expected risk-adjusted profit from the loan. Of course they are buying the loans at a slight discount to what they perceive as actual value (as in nearly any transaction) but that's true of all loans they purchase, regardless of whether the loan's true value is 107 (predatory) or 100.5 (non-predatory). You're not just saying that the pricing favors investors - you're saying that investors make more money from buying high-premium loans than buying low-premium loans. In that case, why wouldn't the market correct this pricing discrepancy?
I'm not sure if I understand your comments about prepayment characteristics either. Yes someone may pay up for a loan because its precise prepayment characteristics, but that's how prices for all loans are determined. I don't see what result in this special extra profit for predatory loans from anyone's perspective but the originator's.
What I never understood is where does the neg-am accrual go?
incessant, remember that you can have "accrual bonds" in the REMIC. This is dicey stuff, because neg am loans involve potential negative amortization, not scheduled neg am. So if the things negatively amortize much faster than you predicted, you can sure have yourself a cash-flow problem.
The 105 in my example is price to the originator of the loan. That means the issuer of the security has to get a sufficiently high return, either in the price of the bond or the yield on its retained piece, to make up for that extra 5 cents it paid for every dollar of principal.
You're not just saying that the pricing favors investors - you're saying that investors make more money from buying high-premium loans than buying low-premium loans. In that case, why wouldn't the market correct this pricing discrepancy?
Bearish, I'm not actually suggesting that any real money is being made off any loan any conduit ever buys for 105. Nonetheless, conduits are out there paying 105; this is a fact.
So you tell me: why is any conduit putting 105 on its rate sheet? How is anybody earning that 5 cents in this day and age with refinance opportunites where they are?
Tanta said: "Sebastian, nobody who fails to stress-test things gets to turn around and excuse himself on the grounds that problems having to do with stresses that had a high probability of happening were "unanticipated" or "unforseen."..."
I agree, but that wasn't what I was asking about. You're "scared" about what you're seeing. Okay, what do you think is going to happen, why, and what would trigger it?
Here's where this question is coming from. CR's estimate of several hundred thousand residential job losses this year (based on significant drops in housing permits, starts, etc.) is looking less and less likely. His reasons (which seemed logical and were well-documented) aren't producing the results he expected.
I'd like to understand your concerns on this other issue so I can make my own appraisal.
felt like back to my mba classroom. We did a lot of prepayment simulations in the class and see how to price under different assumption. However i remember we never discussed about the risk of foreclosure. No one asked why and I guess prof believed the risk of foreclosure was taken care of by lawyers.
but anyway, my first impression is these whole innovations improved market efficiency, and as a result of improved efficiency, risk premium dropped. Is it a good thing? I still believe the blame lies with loan broker/originators for the mess we have now. They have no right to approve a shaky loan no matter how high the demand of secondary market is. Maybe HF or Pension managers didn't fully realize the risk of foreclosure. But at least they put buyback in the contract. What did loan originator do? Approve liar loans, sell them with recourse. Of all people involved, they know the risks of default/foreclosure best, and they choose to look the other way.
i don't want to get into debate of who to blame. there's market, and there's court. let it play out and we see the result soon. let the market make the decision although it might be painful for some. but isn't it part of life?
What did loan originator do? Approve liar loans, sell them with recourse. Of all people involved, they know the risks of default/foreclosure best, and they choose to look the other way.
But you are assuming that the originator made up the underwriting rules. Did you catch the post below on Bear Stearns? That's the securitizer setting what it considers "prudent" underwriting guidelines. And those guidelines are attached to every damned securitization. Why are the ultimate buyers of these securities being given a pass for not having read the prospectus attachments?
Wow! Tanta, you've outdone yourself! I tend to be able to follow your posts for the most part, but it will be quite a while before I have a good understanding of these REMICs. Of course, now I'm more aware of the complexity of the secondary market and I'm more aware of what I don't know. Thank you!
Sebastian: I did not write this post, or any comments thereto, for the purpose of making an argument about job losses or supporting CR's argument about job losses. I thought it would be obvious to one and all why it is frightening to me to contemplate the idea that much money is being funnelled into the mortgage market by people who may not understand what they are or are not getting back for it in terms of risk. But if you are puzzled by that, by all means stay fearless.
What some people here know, that you may not know, is that I haven't even gotten to the whole "credit enhancement" side of the REMIC yet. And that rating agency stuff and who takes credit losses. There's way more complexity left to explore here.
Then after that we'll have to talk about this CDO stuff, I guess, since people keep bringing it up in the comments. I hate to say this, but REMICs are child's play next to CDOs.
I had figured that originaters/loan holders would want to help all the Alt-A/Subprime folks with the upcoming resets refi into something more affordable, afterall, is that not preferable to staright out default? If I understand this correctly it explains why this is not possible without the cash, or equity, to pay the prepay penalties. Since the loan has already been packaged into some kind of security such as these REMICs and if the prepay penalty were to be "excused", then the refi would pretty much be equivalent to a default. As far as the REMIC is concerned since no tranche gets any money. Unless there is a slice of the pie that bets on defaults?
Could loan originaters/REMIC issuers be capable of "gaming" the bond market? Such as pushing large numbers of people into exotic products that will force most of them to refi after the teaser period. Who does oversite on this stuff?
Oh, excellent question. The Fed, OCC, OTS, etc. oversee depository originators. The states--kind of--oversee non-depository originators. The SEC is the one who oversees trading in structured securities. The FTC, and to some extent the Fed, oversee consumer protection issues.
There is no regulator who puts two and two together, as it were.
Given all these tranches, classes, and strips within a single REMIC, how much influence does the loan servicing entity have in determining which classes benefit or lose when individual loans in the pool start to go sour? Are there scenarios where allowing non-paying loan s to float a few more months benefits a strip being paid real cash for computed neg-am, vs. early aggressive foreclosure benefiting a principal-only strip? If so, how much wiggle-room does the servicer have in determining which investors benefit?
I thought this was interesting in the Indymax stuff:
"Indymac has recently been cutting back its product and underwriting guidelines for two reasons: 1) to better align risks and returns in the current environment and 2) due to temporary poor liquidity in the secondary market on certain products. The attached schedules show what the impact of these cutbacks would have been on our overall production for the first quarter. Had the cutbacks been fully in place for the quarter, our overall production would have been reduced by $8.6 billion, or a 33 percent reduction."
six percent, that's part of the whole "class warfare" issue that I hope to follow up on in a future post. Suffice it to say that every decision a servicer takes may benefit one tranche at the expense of another. It's not that the servicer picks who benefits--after all, the loan in question "belongs to" the whole REMIC, not one tranche. But this point does, I think, mean that certain paranoid theories that have been making the rounds about servicers delaying FC in order to "make things look better" are naive. For any tranche that benefitted by that action, another might lose.
Tanta,
Over a large cup of sodium enriched soup, I re-read it and I got it.
There's one term of which I'm not sure of the meaning; Age Adjusted Spread. Is that the estimated present value (in points) of the spread throughout the life of the bond?
I have to admit, I sometimes lose track with the myriad of terms and acronyms. Maybe we should set up a "Calculated Riskionary" or "Encyclopedia Tantanica"?
A record volume of short sales by pessimistic hedge funds and other bearish naysayers is getting absolutely crushed. This market is rising because earnings are once again outperforming consensus estimatesand because interest rates are low, stable, and unlikely to change any time in the foreseeable future.
The mild consumer price index is a big contributor to this outlook of stable interest rates.
A powerful one-two combo of rising profits (which are mothers milk of the stock market) and steady interest rates is a wonderful prescription for soaring stocks and rising American wealth.
Whats that? Did someone say the greatest story never told?
lama, I didn't actually mean Age Adjusted Spread. I meant plain old age adjusted spread. In other words, I don't know if you'd find such a term in any encyclopedia, Tantanic, Britannic, or Satanic.
All I meant was that if the loans in the underlying pool were not all originated in the same month, then they could have a range of interest rates just because the overall market moved during the period they were originated. So, say, "par" is 6.25% in an April coupon, but 6.50% in a May coupon. Both of those borrowers would be paying no points.
But if you had loans that were all, say, three months old when they went into the REMIC, and the note rates ranged by 5 points, that would tell you that borrowers took higher or lower interest rates in the same month, meaning some paid points and some didn't, or some are much worse credit quality than others, or, theoretically, some got gouged and some didn't. You couldn't know which is which without looking at the loans.
Tanta said: "...Sebastian: I did not write this post, or any comments thereto, for the purpose of making an argument about job losses or supporting CR's argument about job losses. I thought it would be obvious to one and all why it is frightening to me to contemplate the idea that much money is being funnelled into the mortgage market by people who may not understand what they are or are not getting back for it in terms of risk. But if you are puzzled by that, by all means stay fearless..."
Tanta, no need for a defensive tone.
I wasn't asking for you to justify CR's forecast, or implying any connection between what he says and what you say.
My point was that CR was seeing "trouble" that so far hasn't amounted to much, even though the evidence he presented appeared to be a compelling case for serious job-loss in residential construction.
You are evidently alarmed by what you're seeing in your area of expertise, but so far it looks as though a group of poorly-capitalized (for the risk they had) lenders has gone bankrupt...and there aren't any serious repercussions developing beyond that. Scary headlines, scary speculation, scary isolated anecdotes of foreclosures, but overall a relatively muted impact on the total housing market.
I'm not at all puzzled by the problems inherent in a group of homebuyers, mortgage brokers, banks, etc., not fully understanding what they're doing. However, financial ignorance and irresponsibility is not a concept that's only just developed in the past few years. (MEW isn't a new concept, either, but that's another tale for another day.)
What I'm asking about is why you think the problem is serious.
My experience lies outside the mortgage markets, but I have lots of experience with failing companies and industries, so perhaps I can answer why we haven't had much damage so far.
From what I gather, the MBS market is teaching everybody the difference between liquidity and solvency (as I like to put it, the difference between not having adequate cash in your wallet vs not having enough cash in the bank)
The foundation of the MBS market is real estate. Prices are undoubtedly falling. This fall has already created a liquidity crisis, which sank some of the undercapitalised players. Liquididty crises tend to blow up companies quickly.
If prices continue to fall, I expect we'll start to see more solvency issues. These sorts of crises tend to happen far more slowly.
Of course, crises have feedback loops where solvency issues create liquidity crises and visa versa.
Excellent summary. I guess Alan Greenspan and the Feds attempt to transfer interest rate risk off of bank balance sheets to consumers has blown up in their faces. The adjustable rate loans that were created have reduced bank's interest rate risk, but increased their credit risk as the adjustable rate loans begin to reprice not withstanding the advanced financial engineering.
Everyone is looking at the MBS market like it was just invented. Historically, and sure it could be different this time, but the securities discussed here didn't have problems from 'credit' but rather interest rates. All of the great MBS explosions (Granite in early 90s) were related to dramatic moves in rates.
"So you tell me: why is any conduit putting 105 on its rate sheet? How is anybody earning that 5 cents in this day and age with refinance opportunites where they are?"
IndyMac released some PR this am that is rich:
Expired
It's as if news of tightening lending standards have been dismissed at NDE with maximum contempt.
Makes all the chatter about bailouts so much more palatable, eh?
I started looking at the NDE PR this morning and had to go get dressed and take a walk.
I really always had some respect for Michael Perry. But this kind of "overdisclosure" that actually conceals "underdisclosure" in a sea of numbers is pissing me off.
I'll try to get a coherent post up on it.
I didn't understand it the first read through. I'll do a REMIC Remix at lunch and see if I can digest it then.
Alas, my Mom never played Mozart for me.
rcyran, start here:
HousingWire || financial news for the mortgage market
Hi Tanta-
seems like some of my comments were cut off. I knew Grant Thornton (of Parmalat fame!) had quit Accredited and Fremont but didn't understand your aside.
I assume it had something to do with either kickbacks or not marking to market properly (or at all). Could you please explain?
By the way, this blog is outstanding.
-Rob
lama, I really really want your final feedback on this. I'm willing to take 100% responsibility for not explaining things well. But I also, assuming that my communication skills aren't that awful, want to know whether a person such as you, with above-average financial skills, can go along with me here at a summary level.
I think this is important for two reasons: one, I'm just talking to myself about some issues if a widely-drawn readership of financially sophisticated people have been so unexposed to the structured MBS market. That may be why my personal skepticism about the extent to which a lot of risks have been truly "dispersed" doesn't make sense to other people.
Two, if someone with a CPA and better than average attention span finds it mind-numbing (and it is), then it's just like conversations about neg am ARMs. The fact that someone is taking one isn't saying much about why someone is taking one. Same goes with the securities we put them in. There are some sharpies buying stuff for hedge purposes, some institutional investors who are just buying a rating, and then some other group of investors who, frankly, may be about as up to speed as the Option ARM borrowers. That idea, if it's true, should frighten everyone. It frightens me.
Tanta - I bet you could put this article to a great Dr. Seuss limerick. It also might make it a tad easier to understand for some of us who are closer to Dr. Seuss level than to yours...
Everything has gotten so amazingly complicated, it can be overwhelming. The thought of escaping to the woods is a temptation, except that the developers already have their eye on that parcel and the property taxes are unaffordable anyway...
Thanks, Rob.
Actually, it was a mere note because I don't know exactly what GT was worried about. But I fear the issue may have something to do with these "held for sale" pipelines.
An outfit like LEND or FMT may do some retail originating, but in most cases they are putting out a price to a correspondent or a broker. What they pay for a loan, obviously, is a question of what they think they'll get for it, either selling it as a whole loan to some other conduit or securitizing it themselves. In other words, there's a "bid-ask" spread in the pricing at this level.
So when I start to see a bunch of parties disclosing some surprising losses on the HFS pipeline, I do not necessarily conclude that the problem is that the market just happened to fall away on these guys and suddenly they can't get the bid they should have gotten on the resale of the loans. It seems possible to me that they just mispriced it up front, counting on the market to keep rising and bailing them out of it. In my experience this happens when someone is desperately chasing market share: it prices some loan at 105 to the originator just to take that loan away from the competition, who is pricing it (perhaps more realistically) at 104.5. You can spend some time doing that if you can keep securitizing what you buy. If the securitization market falls off, and you have to sell it as a whole loan, you get your ass bitten.
This is also a famous trick of outfits (AHM, perhaps?) who have a retail business to support. You have to keep loans coming in because retail is real, true overhead costs. So you get tempted to "subsidize" retail by mispricing. My impression is that that's what happened to one of the first imploders, MLN, although I haven't followed that story fully.
In any case, it can boil down to your auditor forcing you to take impairments on the HFS that you don't want to take.
Tanta, could you expand on the nature of your concern?
Is it that this entire structure of "exotics" has never been under any real stress, so there may be unforeseen and unavoidable consequences?
Sebastia
I thought this was admirably clear, Tanta, and it is a wonderful effort.
Regarding whole loans vs tranched securities, it is worth contemplating that net flow of income to bondholders on tranched securities cannot be more than the net flow on the same whole loan pool. If securitized loans are selling for more overall than whole loans, it is because somewhere, somehow, someone is getting sold a bill of goods.
The essential for securities markets is full disclosure, and one would strongly suspect that full disclosure is impossible for some of these tranches. No one knows how they will perform long term.
This article on REMIC's will take more than a couple of read throughs for me to understand the layering and pricing of these debt/hedge instruments.
But I have a question. Did some of these exotic loans come about because the mortgage refi market for super prime borrowers dry up after the big refi boom during the summer/fall of 2003? Remember that was the time when ten year treasuies hit rates not seen for 40 years. The reason I am asking is because some investor some where, probably in China or Japan, is loosing a ton of money on my 4.36 15 year fixed with 11.5 to go. Having had 14 percent mortgages in the early eighties and having refied many times as the rates ratcheeted down. We like other boomers are now at a point in our lives where you can't and wont move unless downsizing with a complete cash purchase. The inverted yield curve has forced the creation of all these different MBS just to keep a long line of people employed and still in the game. The only problem is that the yield curve was not supposed to stay inverted this long. Certainly this is screwing the models that were used to create these exotic securities.
A confused but curious reader.
Tanta,
You touch here on a question that has interested me for some time.
Who exactly profits from prepayment penalties, and is this likely to affect the courses(s) of action taken in the current time of troubles?
Thanks, MOM. I think it's quite possible that someone is being sold a bill of goods. I just want to say one thing for today: given how much piling on has been going on, generally, with borrowers who took dumb loans (hey! they got what they deserved! they shoulda paid attention!), I am hereby ready to pile on any bozo who bought a tranche of this stuff. Sauce for the goose is sauce for the gander: the whining from bond-land is already starting about how nobody told them that this deal has gigantic risks. So if homeowners have to suck it up, then the rentier class has to suck it up.
Sebastian, nobody who fails to stress-test things gets to turn around and excuse himself on the grounds that problems having to do with stresses that had a high probability of happening were "unanticipated" or "unforseen."
Did some of these exotic loans come about because the mortgage refi market for super prime borrowers dry up after the big refi boom during the summer/fall of 2003?
Short answer is yes. Exactly. And you can plot that (exotic product share/refi share/share of loans in the money for a fixed-rate refi) on a graph. If I had one that was publically available I'd post it. I may have something I can scrape up on that.
Anybody buying fixed rate paper has to have some kind of hedge on it, or else, as you point out, you own an investment yielding in the mid 400bps when new ones are in the mid 600 bps range. Hedges, though, are expensive. As far as I'm concerned, Mr. Greenspan came right out a couple of years ago and said to depositories, put those customers in ARMs so that the customer takes the hedge. Jerk.
ozajh, it's not clear to me what the answer to that is in exact percentages. The investor can take it or the servicer can, and it will depend on the servicing agreement of the REMIC. I'd say this: it used to be investor money (offsetting the yield problem with a prepayment) in the "pass-through" days. More often now it is servicer money (offsetting the servicing fee problem with a prepayment) in the structured security days. I would hazard a guess that in most recent issues, the prepayment penalty goes to the servicer.
I think I just saw the one that got away. Looks as though Mr Seiders and the HB lobby is getting some face time with the Agencies:
"The loans Freddie Mac buys under this program would not be limited to refinancing, though refinancing is the initial focus now that millions of people have adjustable-rate mortgages with low teaser rates that will soon spike."
So, looks like they are throwing in a little something for the HBs too. I was puzzled at why the HB stocks have gotten a bounce in the last few days.
The market economy, where no well-heeled, sharply directed influence goes unrewarded. The gravy trains keep running. They just never seem to stop and pick me up.
Nice article, Tanta! Thanks for taking the time to educate this one member of the unwashed masses.
I was taken in by the IndyMac article, however; in particular, how factors in the S&P LEVELS(tm) model are literally driving the changes. Leave aside the fact that a regression model based upon backwards-looking historical data is likely to be stunningly less predictive when faced with novel mortgage vehicles with built-in discontinuities. What took me was their faith-based belief that the risk factors were independent, and that tweaking the weights of some parameters of the model would suffice to control other risk factors:
"Lower documentation requirements on loans are more than offset by other compensating factors, such as lower CLTVs and higher FICOs, with the result that lower documentation loans actually have lower estimated lifetime losses."
Another interpretation is that the model underweights the risk of certain documentation types due to lack of historical precedent, making it cheap to offset the effect of that risk in the model by tightening up on an overweighted factor. I hope these securitizer using this LEVELS(tm) model are pricing in "model risk": the likelihood that the next few years data will lead to a greatly-changed LEVELSPLUS(tm) model that accounts for all the soon-to-be-noticed failures of the current scheme.
Tanta,
Very interesting. As background, M.S. in Aerospace Engineering. I think I understand the jist, but I need to make a diagram and some equations to see how this stuff works practically.
What I never understood is where does the neg-am accrual go? They can't be paying money out to the bondholders that doesn't exist, right? Does that mean somebody has signed up to get neg-am payments, and in the interim, they get the interest paid on the neg-am portion each period? Where does the slush fund live? Does the initial premium paid (the 5 in "105 cents on the dollar") give the funding to keep the cash flow from cutting off in the early months?
Until I read Tanta's lucid description of the mortgage repackaging market, I had always felt that there was a law of conservation of complexity in financial markets.
Very complex markets often develop around widely held simple tradables (like shares for example).
On the other end, you get simple markets for really complex things. Considering something complex like power plants for example, there are very few buyers and sellers so you don't really need a sophisticated marketplace.
The mortgage market is the first example that I have seen of a complex marketplace for a complex commodity.
DCRogers, you are so right. I think it's even worse than that, because I don't trust their "CLTV" numbers. I'll try to post something on that later today.
Tanta wrote: "The claim that premium pricing prevents rate predationthat the investor doesnt gain by having a borrower pay an above-market rate because it pays too much premium for the loanis a mite disingenuous in the context of structured securities. If you are buying whole loans for an investment portfolio, or for inclusion in a simple pass-through, that might make some sense. But if you are buying loans for a security in which someone will benefit from the fast prepayment of that loanand someone else will benefit if it stays on the booksyou may well be paying up for that loan precisely because it has the payment and prepayment characteristics you want, not in spite of them. There is always someone on the other side of a hedge trade. If there were no incentive for conduits to pay 105 for a loan, I can assure you that they wouldnt pay it, or not for long."
The claim is that the rate the borrower pays is irrelevant for the wall street conduit or any other investor, in terms of expected risk-adjusted profit from the loan. Of course they are buying the loans at a slight discount to what they perceive as actual value (as in nearly any transaction) but that's true of all loans they purchase, regardless of whether the loan's true value is 107 (predatory) or 100.5 (non-predatory). You're not just saying that the pricing favors investors - you're saying that investors make more money from buying high-premium loans than buying low-premium loans. In that case, why wouldn't the market correct this pricing discrepancy?
I'm not sure if I understand your comments about prepayment characteristics either. Yes someone may pay up for a loan because its precise prepayment characteristics, but that's how prices for all loans are determined. I don't see what result in this special extra profit for predatory loans from anyone's perspective but the originator's.
What I never understood is where does the neg-am accrual go?
incessant, remember that you can have "accrual bonds" in the REMIC. This is dicey stuff, because neg am loans involve potential negative amortization, not scheduled neg am. So if the things negatively amortize much faster than you predicted, you can sure have yourself a cash-flow problem.
The 105 in my example is price to the originator of the loan. That means the issuer of the security has to get a sufficiently high return, either in the price of the bond or the yield on its retained piece, to make up for that extra 5 cents it paid for every dollar of principal.
The mortgage market is the first example that I have seen of a complex marketplace for a complex commodity.
What about the international currency markets?
Tanta
Thanks for the elightenment but I'm afraid you crossed my eyes on this one. Will spend the wknd trying to disgest it.
You're not just saying that the pricing favors investors - you're saying that investors make more money from buying high-premium loans than buying low-premium loans. In that case, why wouldn't the market correct this pricing discrepancy?
Bearish, I'm not actually suggesting that any real money is being made off any loan any conduit ever buys for 105. Nonetheless, conduits are out there paying 105; this is a fact.
So you tell me: why is any conduit putting 105 on its rate sheet? How is anybody earning that 5 cents in this day and age with refinance opportunites where they are?
Tanta said: "Sebastian, nobody who fails to stress-test things gets to turn around and excuse himself on the grounds that problems having to do with stresses that had a high probability of happening were "unanticipated" or "unforseen."..."
I agree, but that wasn't what I was asking about.
You're "scared" about what you're seeing. Okay, what do you think is going to happen, why, and what would trigger it?
Here's where this question is coming from. CR's estimate of several hundred thousand residential job losses this year (based on significant drops in housing permits, starts, etc.) is looking less and less likely. His reasons (which seemed logical and were well-documented) aren't producing the results he expected.
I'd like to understand your concerns on this other issue so I can make my own appraisal.
Sebastia
felt like back to my mba classroom. We did a lot of prepayment simulations in the class and see how to price under different assumption. However i remember we never discussed about the risk of foreclosure. No one asked why and I guess prof believed the risk of foreclosure was taken care of by lawyers.
but anyway, my first impression is these whole innovations improved market efficiency, and as a result of improved efficiency, risk premium dropped. Is it a good thing? I still believe the blame lies with loan broker/originators for the mess we have now. They have no right to approve a shaky loan no matter how high the demand of secondary market is. Maybe HF or Pension managers didn't fully realize the risk of foreclosure. But at least they put buyback in the contract. What did loan originator do? Approve liar loans, sell them with recourse. Of all people involved, they know the risks of default/foreclosure best, and they choose to look the other way.
i don't want to get into debate of who to blame. there's market, and there's court. let it play out and we see the result soon. let the market make the decision although it might be painful for some. but isn't it part of life?
What did loan originator do? Approve liar loans, sell them with recourse. Of all people involved, they know the risks of default/foreclosure best, and they choose to look the other way.
But you are assuming that the originator made up the underwriting rules. Did you catch the post below on Bear Stearns? That's the securitizer setting what it considers "prudent" underwriting guidelines. And those guidelines are attached to every damned securitization. Why are the ultimate buyers of these securities being given a pass for not having read the prospectus attachments?
Wow! Tanta, you've outdone yourself! I tend to be able to follow your posts for the most part, but it will be quite a while before I have a good understanding of these REMICs. Of course, now I'm more aware of the complexity of the secondary market and I'm more aware of what I don't know. Thank you!
Sebastian: I did not write this post, or any comments thereto, for the purpose of making an argument about job losses or supporting CR's argument about job losses. I thought it would be obvious to one and all why it is frightening to me to contemplate the idea that much money is being funnelled into the mortgage market by people who may not understand what they are or are not getting back for it in terms of risk. But if you are puzzled by that, by all means stay fearless.
Thanks, Todd.
What some people here know, that you may not know, is that I haven't even gotten to the whole "credit enhancement" side of the REMIC yet. And that rating agency stuff and who takes credit losses. There's way more complexity left to explore here.
Then after that we'll have to talk about this CDO stuff, I guess, since people keep bringing it up in the comments. I hate to say this, but REMICs are child's play next to CDOs.
I could stop now while we're all still innocent.
I had figured that originaters/loan holders would want to help all the Alt-A/Subprime folks with the upcoming resets refi into something more affordable, afterall, is that not preferable to staright out default? If I understand this correctly it explains why this is not possible without the cash, or equity, to pay the prepay penalties. Since the loan has already been packaged into some kind of security such as these REMICs and if the prepay penalty were to be "excused", then the refi would pretty much be equivalent to a default. As far as the REMIC is concerned since no tranche gets any money. Unless there is a slice of the pie that bets on defaults?
Could loan originaters/REMIC issuers be capable of "gaming" the bond market? Such as pushing large numbers of people into exotic products that will force most of them to refi after the teaser period. Who does oversite on this stuff?
Who does oversite on this stuff?
Oh, excellent question. The Fed, OCC, OTS, etc. oversee depository originators. The states--kind of--oversee non-depository originators. The SEC is the one who oversees trading in structured securities. The FTC, and to some extent the Fed, oversee consumer protection issues.
There is no regulator who puts two and two together, as it were.
Given all these tranches, classes, and strips within a single REMIC, how much influence does the loan servicing entity have in determining which classes benefit or lose when individual loans in the pool start to go sour? Are there scenarios where allowing non-paying loan s to float a few more months benefits a strip being paid real cash for computed neg-am, vs. early aggressive foreclosure benefiting a principal-only strip? If so, how much wiggle-room does the servicer have in determining which investors benefit?
H&R Block unloading Option One to Cerberus, after an additional 40% haircut:
H&R Block to Sell Option One Mortgage to Cerberus (Update5) - Bloomberg.com
Everything has a price!
I thought this was interesting in the Indymax stuff:
"Indymac has recently been cutting back its product and underwriting guidelines for two reasons: 1) to better align risks and returns in the current environment and 2) due to temporary poor liquidity in the secondary market on certain products. The attached schedules show what the impact of these cutbacks would have been on our overall production for the first quarter. Had the cutbacks been fully in place for the quarter, our overall production would have been reduced by $8.6 billion, or a 33 percent reduction."
six percent, that's part of the whole "class warfare" issue that I hope to follow up on in a future post. Suffice it to say that every decision a servicer takes may benefit one tranche at the expense of another. It's not that the servicer picks who benefits--after all, the loan in question "belongs to" the whole REMIC, not one tranche. But this point does, I think, mean that certain paranoid theories that have been making the rounds about servicers delaying FC in order to "make things look better" are naive. For any tranche that benefitted by that action, another might lose.
Tanta,
Over a large cup of sodium enriched soup, I re-read it and I got it.
There's one term of which I'm not sure of the meaning; Age Adjusted Spread. Is that the estimated present value (in points) of the spread throughout the life of the bond?
I have to admit, I sometimes lose track with the myriad of terms and acronyms. Maybe we should set up a "Calculated Riskionary" or "Encyclopedia Tantanica"?
ATTENTION ANONY BLOGGING SCARED BEARS:
The Goldilocks stock market is booming.
A record volume of short sales by pessimistic hedge funds and other bearish naysayers is getting absolutely crushed. This market is rising because earnings are once again outperforming consensus estimatesand because interest rates are low, stable, and unlikely to change any time in the foreseeable future.
The mild consumer price index is a big contributor to this outlook of stable interest rates.
A powerful one-two combo of rising profits (which are mothers milk of the stock market) and steady interest rates is a wonderful prescription for soaring stocks and rising American wealth.
Whats that? Did someone say the greatest story never told?
PULL MY FINGER...........AAAAAHHHHHHHHHHH!
As long as profits are rising, it must be all good!
lama, I didn't actually mean Age Adjusted Spread. I meant plain old age adjusted spread. In other words, I don't know if you'd find such a term in any encyclopedia, Tantanic, Britannic, or Satanic.
All I meant was that if the loans in the underlying pool were not all originated in the same month, then they could have a range of interest rates just because the overall market moved during the period they were originated. So, say, "par" is 6.25% in an April coupon, but 6.50% in a May coupon. Both of those borrowers would be paying no points.
But if you had loans that were all, say, three months old when they went into the REMIC, and the note rates ranged by 5 points, that would tell you that borrowers took higher or lower interest rates in the same month, meaning some paid points and some didn't, or some are much worse credit quality than others, or, theoretically, some got gouged and some didn't. You couldn't know which is which without looking at the loans.
ozajh, you can also create prepayment penalty securities. Generally the residual owner also holds this.
Thanks for these educational (and entertaining) explanations, Tanta.
Tanta said: "...Sebastian: I did not write this post, or any comments thereto, for the purpose of making an argument about job losses or supporting CR's argument about job losses. I thought it would be obvious to one and all why it is frightening to me to contemplate the idea that much money is being funnelled into the mortgage market by people who may not understand what they are or are not getting back for it in terms of risk. But if you are puzzled by that, by all means stay fearless..."
Tanta, no need for a defensive tone.
I wasn't asking for you to justify CR's forecast, or implying any connection between what he says and what you say.
My point was that CR was seeing "trouble" that so far hasn't amounted to much, even though the evidence he presented appeared to be a compelling case for serious job-loss in residential construction.
You are evidently alarmed by what you're seeing in your area of expertise, but so far it looks as though a group of poorly-capitalized (for the risk they had) lenders has gone bankrupt...and there aren't any serious repercussions developing beyond that. Scary headlines, scary speculation, scary isolated anecdotes of foreclosures, but overall a relatively muted impact on the total housing market.
I'm not at all puzzled by the problems inherent in a group of homebuyers, mortgage brokers, banks, etc., not fully understanding what they're doing. However, financial ignorance and irresponsibility is not a concept that's only just developed in the past few years.
(MEW isn't a new concept, either, but that's another tale for another day.)
What I'm asking about is why you think the problem is serious.
Sebastia
Sebastian-
My experience lies outside the mortgage markets, but I have lots of experience with failing companies and industries, so perhaps I can answer why we haven't had much damage so far.
From what I gather, the MBS market is teaching everybody the difference between liquidity and solvency (as I like to put it, the difference between not having adequate cash in your wallet vs not having enough cash in the bank)
The foundation of the MBS market is real estate. Prices are undoubtedly falling. This fall has already created a liquidity crisis, which sank some of the undercapitalised players. Liquididty crises tend to blow up companies quickly.
If prices continue to fall, I expect we'll start to see more solvency issues. These sorts of crises tend to happen far more slowly.
Of course, crises have feedback loops where solvency issues create liquidity crises and visa versa.
Excellent summary. I guess Alan Greenspan and the Feds attempt to transfer interest rate risk off of bank balance sheets to consumers has blown up in their faces. The adjustable rate loans that were created have reduced bank's interest rate risk, but increased their credit risk as the adjustable rate loans begin to reprice not withstanding the advanced financial engineering.
Credit derivatives
At the risky end of finance
Premium content | Economist.com
Everyone is looking at the MBS market like it was just invented. Historically, and sure it could be different this time, but the securities discussed here didn't have problems from 'credit' but rather interest rates. All of the great MBS explosions (Granite in early 90s) were related to dramatic moves in rates.
"So you tell me: why is any conduit putting 105 on its rate sheet? How is anybody earning that 5 cents in this day and age with refinance opportunites where they are?"
PP
Having fun yet? - Absolutely! It occurs to me that when you're finished with the UberNerd series you will have the first draft of a very nice book.
Tanta,
I think a good presentation that links your REMIC discussion with tranched CDOs of ABS and the potential for trouble is shown here...
http://www.hudson.org/files/Publications/YangComments.ppt
Tanta/Bacon Dreamz,
Thanks for the replies.
"What I'm asking about is why you think the problem is serious."
Hey Sebastian! It's serious!!