« MR BookForum | Main | He didn't even mention bottleneck guitar »
Bernard Guerrero, from the comments
Sub-prime delinquency has increased drastically in period-to-period % terms, but from an unnaturally low base. What has driven the current crisis is the extreme effective leverage applied by some of the ABS buyers. You can make a AAA tranche out of sub-prime paper, but only by putting it in a last-loss position vis-a-vis a first-loss "equity" tranche. Buy that equity tranche and you've already applied leverage. Buy it on deep margin and you've got a lever big enough to shift the Moon. And the slightest bump wipes you out. This has very little to do with work-out difficulties (and work-outs often don't, um, work out anyway) or less info on the part of new-style lenders (they have less info, to be sure, but it shows in pricing differentials; I regularly see us undercut by credit unions that know their customers well). It has to do with the risk that must exist if you're getting outsized returns, even if you don't see it.
Here is the original post.
Posted by Tyler Cowen on August 17, 2007 at 01:50 PM in Economics | Permalink
Comments
What does that mean?
Posted by: Bergamot at Aug 17, 2007 2:29:22 PM
Conspicuous vocabularic waste.
Posted by: thorstein veblen at Aug 17, 2007 2:36:55 PM
Funny, I had this exact conversation over lunch with a colleague. For our clients (large insurance companies) the risk profile of the AAA tranches they hold aren't drastically different, it's only when you move down the ladder that it becomes an issue. It's the mispricing (or mis-rating in hindsight) of the lower tranches that funds will get killed on, and frankly, there aren't too many widows and orphans out there picking up z-tranches. Should funds get bailed out for taking these positions? For one, it's favouring short term liquidity over long-term liquidity in the sense that a bailout today will, while helping clear the logjam today, distort the long-term valuations of these bonds away from the fundamentals. Second, there's a fund on either side of each transaction, so for every fund that's getting killed today, someone else read it correctly. The downside is that people aren't packaging new bonds until they can price them accurately. Is this a negative turn of events, certainly, but is it a systemic crisis, I think not.
Posted by: Aschkan at Aug 17, 2007 2:47:00 PM
Well, I do like seeing myself talk, so to speak. However, assuming that Bergamot's "what does that mean" is referring to the tranches, I'll expand. To simplify a great deal, if you have some underlying asset class with, say, a 10% PDF (or default frequency), you can still make a AAA tranche out of it by putting some bonds based on it in a last-loss position and then making a much riskier first-loss "equity" tranche. The equity side gets paid hugely if the loss never materializes, but gets crushed if the expected 10% losses or worse show up. The AAA tranche gets paid less in yield terms, but always has first dibs on the actual cash-flow that comes in.
Posted by: Bernard Guerrero at Aug 17, 2007 2:48:45 PM
May I ask related questions.
How common is the "slicing" of sub-prime loans? The vast majority? Half? "Some?"
How many slices would be typical? Two? Five? Ten?
Does anyone anywhere have answers to such matters?
Such facts would seem to have been germane to the rating agencies in the initial securitization. An "unsliced" loan would presumably be more valuable now, though obviously not in the past.
•••
(Btw, this is the first place I have run across the concept of and issues with "re-assembly" of sliced-up loans -- which vastly complicates work-outs and maybe that is all that Krugman meant -- and I want to compliment the hosts for creating such a stimulating venue.)
Posted by: David Sucher at Aug 17, 2007 3:32:44 PM
David:
A while back, I put a example of slicing and dicing here:
http://www.aleablog.com/2007/08/06/misleading-index-of-the-year-abxhe/
usually the equity tranche is not sold to the public and it takes 3/4% loss of principal before the lower tranches are impaired.
Posted by: jck at Aug 17, 2007 3:49:03 PM
Bernard makes a nice point, but I think he's missing some other important points. It appears that a lot of the models that informed these debt ratings had overly optimistc assumptions about housing prices.
If you have housing price declines (which we have had and will continue to have), then you start to have problems even up into the AA tranches.
In other words, it isn't just the "high-risk" tranches that have problems, although admittedly their problems are far more spectacular.
Posted by: Keith at Aug 17, 2007 5:28:12 PM
Keith,
I'll agree with you and go a bit further.
A) I believe LTV has a direct impact on likelihood of consumer default, all else being equal. We accept that a corporation will, in theory, default once it's assets are worth less than it's debts (see Merton), or in other words, that the equity is gone. The average consumer may not operate with such crystalline clarity, but they clearly take the degree to which they are "under water" into account in deciding whether to let the house go back to the bank. (There are reputational costs in defaulting and/or going BK, of course, but let's put that to the side for the moment.)
B) Overcapacity on the producer side led to deteriorating underwriting standards and, by and by, fed the mis-estimation of risk. "Alt-A" paper can't really be prime, for instance. You don't know what the customer makes and yet the risks were treated as largely equivalent. (I suspect that there was selection bias hitting the modeling process, too. If you used to do few alt-A's and they performed well because your human underwriters were stingy with them, the models will tend to reflect better alt-A performance than would be true if you opened the flood-gates.)
All that aside, though, most of the real damage we've seen is associated with either heavily leveraged buyers of the riskiest assets or guys being killed by the illiquid nature of even the less risky assets in a panic environment.
Posted by: Bernard Guerrero at Aug 17, 2007 5:56:17 PM
Securitization has always seemed like black magic to me because by artfully slicing and dicing the risk tranches you can convince yourself there's a triple-A pony in the junkpile (I'm sanitizing the language here). And of course lower interest rates mean higher NPV's and higher asset valuations.
But the lever on the way up is the screw on the way down. I wonder how much weight the rating agencies and Wall Street quants give to the multiplier effect of defaults and rising credit spreads as the asset balloon deflates, thereby increasing the stress on the higher-grade tranches? I guess dynamic analysis only applies to tax policy, not the debt markets.
Posted by: Steve Yuen at Aug 17, 2007 8:41:49 PM
The average consumer may not operate with such crystalline clarity, but they clearly take the degree to which they are "under water" into account in deciding whether to let the house go back to the bank. (There are reputational costs in defaulting and/or going BK, of course, but let's put that to the side for the moment.)
Can we really put these reputational costs to the side? I would say that they are very big factors in most peoples' minds, and disregarding them is very unrealistic.
Posted by: Peter at Aug 17, 2007 9:28:01 PM
Can we really put these reputational costs to the side? I would say that they are very big factors in most peoples' minds, and disregarding them is very unrealistic.
I would suggest that the reputational costs would be one of the things feeding the "we buy ugly houses" industry. They may actually be buffering the entire downturn, altho in my case they only offered 75 for a house which brought me 137.
Posted by: triticale at Aug 17, 2007 11:08:21 PM
[R]isk ... must exist if you're getting outsized returns, even if you don't see it.
It was worth reading that entire passage for this one statement. In fact, it should be taped to everyone's mirror. A considerable percentage of all loss comes from overlooking this.
BTW, in my previous comment, I omitted the fact that I spoke to the "cash buyer" out of curiosity and convenience; I was not quite that desperate to sell. We owned it free and clear but were heading toward property tax issues.
Posted by: triticale at Aug 17, 2007 11:15:22 PM
> How common is the "slicing" of sub-prime loans? The vast majority? Half? "Some?"
1. It is not the loan that is sliced -- it is the pool of several loans. Hopefully this clarifies Bernard's comments aout first loss and last loss to the uninitiated...
2. Yes, it is the vast majority of sub-prime mortgages that is bundled into pools and sliced.
3. Note that the individual loan is usually serviced by the originator of the loan and their servicing vendor. So if there is a default, a work out will happen regardless of the complexity of the securitized "slices".
4. At the time of rating (and of course buying), it was already assumed that a certain percentage will default, and already assumed that there may be a decline in the value of the asset. The first-loss buyers (and perhaps second- and third-loss buyers) took those risks. I do not believe any AAA tranche (last-loss positions) has ever defaulted in the history of securitization. Which is why the Fed has a window which offers to buy them.
Posted by: KP at Aug 18, 2007 12:37:24 PM
I find this a very interesting discussion, but I lack the basic understanding of structured finance to follow the nuances. Can anyone suggest a good primer on how this kind of risk pooling works?
Posted by: mdc at Aug 18, 2007 5:00:20 PM
Two recent good books :
1.Credit Derivatives by Erik Banks , Morton Glantz and Paul Siegel
Worth reading for chapter 3 , structured and synthetic credit products , best introduction to CDOs.
2.Credit Derivatives: A Primer on Credit Risk, Modeling, and Instruments by George Chacko , Anders Sjöman , Hideto Motohashi and Vincent Dessain
Does exactly what it says on the tin :no credit derivatives experience necessary.
Posted by: jck at Aug 18, 2007 5:23:49 PM
To expand on KP's comment, it isn't just pools of sub-prime mortgages that get bundled and securitized, the quality of the assets can be prime and the collateral can be nearly anything that produces a predictable cashflow: auto loans, student loans, credit card debt, etc. And here's another interesting discussion on the topic.
http://accruedint.blogspot.com/2007/07/dangerous-to-your-hedge-fund-commander.html
Posted by: Bernard Guerrero at Aug 18, 2007 7:36:32 PM
btw, where is "Private Mortgage Insurance" in this whole story?
I was under the impression that any loan greater than 80% LTV required it. Did that disappear under Bush?
Posted by: David Sucher at Aug 18, 2007 8:30:02 PM
Aschkan “but is it a systemic crisis, I think not.”
I am not going to enter in some nitpicking on the trenching part of it all but let us be clear that when some lousy awarded mortgages (forget about calling them subprime, they are much worse than that) get catapulted into a global financial problem because some credit rating agencies who were implicitly and explicitly appointed by the regulators get it wrong, we are looking into the land of the systemic risks.
The real truth we have to face is that the better the credit rating agencies get at what they are supposed to do, the larger the build up of really dangerous systemic risks and the bigger the ensuing explosion.
I am not against credit rating agencies. I will use them. But please let us unshackle the markets from having to use them.
Posted by: Per Kurowski at Aug 19, 2007 12:02:51 PM
CDOs are a very interesting debt instrument. I lack complete understanding of how they are administered, but is the reason someone buys into a CCC tranche on margin because the fees of entering the tranche are insanely low? Did the industry take a hit because the credit raters did such a poor job of assessing these CDOs? Do I understand correctly that the senior tranche is the first to gain and the last to lose, but they have to pay very very high fees to sustain the CDO? And lastly, since the lower tranches are not investment grade derivatives, who in god's name buys them?
Another interesting thing that has entered the market fairly recently are Cat Bond CDOs.
Posted by: Michael at Aug 19, 2007 8:50:49 PM
>And lastly, since the lower tranches are not investment grade derivatives, who in god's name buys them?
Very often, insurance companies. Think aout it: buy 10,000 such low grade bonds fo pennies on the dollar, assume 75% loss, and you can still make a bundle...
Posted by: KP at Aug 20, 2007 11:53:11 AM
To everyone wondering about the voodoo, it's really not all that different from a portfolio of risky stocks being combined into a less risky whole. Correllation estimates were much too low (so the risk level was much higher than the rating agencies believed and essentially they decided how large the lower grade tranches had to be to protect the senior parts).
Also, to the person who asked how big these were, they are much larger than the number of loans made through the "miracle" of the synthetic CDO (securitized obligation). The long and short is you use default insurance (known as a credit default swap) to add credit risk (writing insurance) and package the premiums along with the cash flows of a small portion of some prime paper (like treasuries) to create a funded CDO and an unfunded credit spread.
Posted by: nelsonal at Aug 20, 2007 1:35:47 PM
Hi Best wishes。Allow me to offer my heartiest wishes.xicao loves-流水线娱乐博客常年提供高、中、低压锅炉钢管、流体钢管、结构钢管、化肥专用钢管、石油裂化钢管、地质钢管、液压支柱钢管及合金钢管-无缝管-无缝钢管等论文发表资讯/刊物信息，协助客户制定论文发表方案
Posted by: 无缝管 at Nov 13, 2007 8:14:09 PM
power leveling
powerleveling
world of warcraft power leveling
ffxi power leveling
ffxi powerleveling
lotro powerleveling
lotro power leveling
power leveling
powerleveling
world of warcraft power leveling
ffxi power leveling
ffxi powerleveling
ffxi
ffxi gil
lotro powerleveling
lotro power leveling
lotro gold
power leveling
powerleveling
world of warcraft power leveling
powerleveln
power leveln
world of warcraft power leveln
world of warcraft powerleveln
ffxi powerleveling
coh powerleveling
cov power leveling
Posted by: world of warcraft at Nov 26, 2007 11:33:46 PM
China Tours and Yangtze Cruises A professional China travel agency offering private & group package tours, tailor-made vacations for families, couples and individual travelers, Yangtze cruises, hotel & ticket booking services.
China Travel and Tours One-stop China travel service delivering China package tours, discount Yangtze River cruises and best rate hotels.
Tibet Train Travel A reliable tour operator specializing in Tibet tour packages, adventures, hotel and train ticket booking.
China Maps A free web-based source for China travel and tour map users
Voyage Chine, Tour de Chine, Circuit Chine - Un agence spécialiste voyage en Chine
Great Wall of China A website for the Great Wall of China.
Posted by: China tour at Nov 27, 2007 2:38:11 AM
Posted by: 翻译公司 at Feb 13, 2008 10:10:50 AM
