Rug Burn Junky on 26/3/2009 at 22:00
Morte,
I don't know if I can really explain it any better than I already have.
The idea that everyone keeps latching on to: "You can't bundle shit loans and make them AAA" is an overreaction based on failing to understand the investments. It makes for good soundbites, because it seems like common sense, so that's what people walk away with, but the underlying theory is pretty sound.
If you take a sufficiently large pool, there's a virtual certainty that a smaller subset will be able to pay you back. To make up for the ones that don't, you increase the interest on all of them. If you balance it right, the increase in interest is going to be greater than the number of defaults.
You can plug it into a spreadsheet and see. It's a pretty simple formula to model - multiply the original principal by both the rate of defaults and the interest rate, and see just how many defaults you need before you're pulling in less than 10% of your original stake. It's over 90%. It's not reliant upon housing prices continuing to go up, and works just fine if house prices stay flat. This is not some super secret smirking formula, it's basic finance.
The problem is underestimating the default rate due to the super high correlation, and the drastic decline that we're in. It's a real problem, and it needs to be addressed, and the economics of the deals need to be tweaked, but ultimately, yeah, you can wring less risk out of risky loans by bundling them in such a fashion.
Rating Agencies really did drop the ball, and their models need an overhaul, but CDS, or bond insurance, generally work. The problem, again, is correlation. It's like insuring only one town against hurricanes - you can't count on the premiums from everywhere else to carry you through. But requiring less collateral from a company with a strong credit rating is exactly how it works for regular people - at least, until the housing bubble when everyone was competing for borrowers.
As for the CDS, you're mostly on the right track, but really, they do reduce risk as well. Naked CDS are less necessary, but they ultimately they do serve a purpose that is beyond simply betting. I can't model it very easily over a forum, but you can drop them into your portfolio to create synthetic positions that offset other positions, in order to mimic the exposure of buying a certain bond. That actually creates a more smoothly functioning market by eliminating artificial scarcity of a bond and allowing price discovery. That's a tool that should be in the arsenal if you're creating any sort of sophisticated institutional portfolio.
And yeah, I've read Taleb, I love his writing, and he's exactly the style of investor that I wish I were. I read Fooled By Randomness when it came out in paperback, and I've had a copy of his technical book on options for years. I'm fully on board with him but what he's saying is substantially different than your blanket dismissal.
Scots Taffer on 26/3/2009 at 23:28
guess you're over that block, huh
will read in depth at some stage
Tocky on 27/3/2009 at 04:47
Quote:
at least, until the housing bubble when everyone was competing for borrowers.
There were still a few crackheads with no job who wanted a million dollar home out there. Countrywide just didn't give thier loan guys enough incentive to find them.
the_grip on 27/3/2009 at 11:46
Quote Posted by Rug Burn Junky
But this little canard about just trading contra Cramer has been going around for a while, and it's hyperbole - exaggerating both his effect on the market, and other people's responses thereto...
That's not what Cramer always recommends though, and why you can't just point to a stock that was X value one day, and (1/10)X six months later and say “See! He was wrong!” especially if that stock was at 2X at some point in between. A lot of his picks are based on short term technical analysis. If he highlights a stock that has significant upward momentum, you can easily pull up the analyses and see whether there's a legit trade for yourself - Is it trending up? Can you rely on that trend? Is there a good price point to get in? The real pros are making their trades on this basis, ignoring Cramer's analyses entirely, but that doesn't mean he can't unearth something that makes them or anyone else consider a trade in the first place, but ultimately, he's illustrating aspects of trading, not giving instructions to trade, and a professional trader isn't going to need that illustration. So yeah, he's not offering strictly serious useful commentary, but that's pretty obvious from his show. He's a clown.
The "short Cramer picks" is definitely an obsolete method now. That was more pre-October 2007. I think by and large Cramer is ignored in a large part these days.
In my opinion, like you said, he is a clown, and I think his TA is worth less than the time it takes to listen to him dance around and push buttons. But that is just me... I use TA almost exclusively to trade stocks and futures, and I find Cramer more like talk show dribble. He has, as you said, unearthed decent choices from time to time, but I watch him so rarely these days (maybe once a year except on YouTube) that I wouldn't know anymore. If somebody wants to get stock picks (which as you said is a dangerous way to choose stocks), then they are better off finding a good subscription service.
Tocky on 27/3/2009 at 12:14
And here I thought he just bought Jones soda because he could put moms picture on it and then tasted it one day and ran for the phone to sell.
Anyway, I was just wondering how many AIG execs would have gotten bonuses had it been allowed to go under. Under ordinary conditions I would say honor the contracts but getting a bailout to stay afloat is not an ordinary condition. Those execs going to leave are going to anyway to start those smaller companies which will compete with AIG and giving them more now will just help them get started. Or am I wrong about people? I know jack about finance obviously.
Morte on 27/3/2009 at 13:01
RBJ, sorry, I'm mixing the technical merits of a CDO with what seems to have been done with them. I do understand that the AAA tranche has it very easy to get their money back, as long as the fundamentals aren't completely rotten.
What appears to have occurred in quite a few cases though is people creating them out of mortgages that would be worthless unless people could flip their houses or refinance in a rising market, and then waltzing by ratings agencies that were asleep on the job.
That's a clear abuse and/or regulation failure though, and different from whether a CDO is a good and useful financial instrument. On that part of the equation, it seems to me that it'd be very sensitive to modelling errors when judging risks. Not necessarily the AAA tranches, but the mezzanine tranches. And if there were a hundred shorts the mezzanine tranches in the form of CDS, and those were bundled together in a CDO, you'd end up with something that's even more vulnerable.
If you take a hundred loans and assume for argument's sake they've a binomial distribution, and assume you'd need 90% of people to default for the AAA people not to get their money back, and 15% for the mezzanine tranch. Then do the same for a hundred CDS shorts on the mezzanines. Say you have a fairly horrendous 10% default rate, and things still look pretty rosy. There's only a 7% chance the mezzanine tranch won't get their money back, and for the CDS-based tranch, it's only a 1% chance. Bump the default rate to 11% though, and it gets to 13% and 35% percent respectively. With a 12% default rate, it's 22% and 96%. 20% default rate, and it's 92% for the original, and a 100% chance for the mezzanine and an 82% chance for the supposedly secure AAA tranch to lose their money in the CDS derived CDO.
It's an extremely naive example, but it does illustrate that the more complex you make the instruments, the more you'll expose yourself to flaws in the models. I'm not saying the CDO has to be thrown out completely, but from my layman's corner I'd like to see some pretty hard limits on how much exposure to them any one company can have, and what sort of things can be put into one. People don't have enough of a handle on them to be given free reign.
Rug Burn Junky on 27/3/2009 at 14:11
Quote Posted by the_grip
The "short Cramer picks" is definitely an obsolete method now.
That's just it, it was never really reliable. You can cherry-pick results to make him look good or bad (and with everything across the board down now, it's pretty easy to make him look bad), but it was never that simple, and it's a very lazy criticism. Anytime you predict the future, you're going to get some calls wrong.
Quote Posted by Tocky
Anyway, I was just wondering how many AIG execs would have gotten bonuses had it been allowed to go under. Under ordinary conditions I would say honor the contracts but getting a bailout to stay afloat is not an ordinary condition.
Sure, they wouldn't have gotten bonuses if the company had been allowed to go under, but after the bailouts took place, many of them would have left much much sooner if they hadn't been promised the money. What happened in many instances was "Look, if you stay on, we promise to pay you $X," they continued to work based on that promise, and then when time came to pay, everyone got up in arms, and said "Fuck you, we ain't giving you jack shit." Which would have been fine if it were said 6 months ago, but once you get people to perform their end of the bargain, it's pretty shitty to turn around and yank the rug from under them, and that's the inequitable part that I have a problem with.
Quote Posted by Morte
RBJ, sorry, I'm mixing the technical merits of a CDO with what seems to have been done with them.
That was really where you were running into problems. That said, and without checking your math explicitly, you're on the right track. It's still a little bit of an overreaction, but I agree in principle.
I will note that everyone -- the bankers, the rating agencies, the insurers, the (ahem) lawyers -- were quite aware of the sensitivity to defaults, and they were heavily disclosed in the offering documents (I should know, I've written them myself, at least, on the underlying MBS deals that operate under the same principle). It's easy to throw rocks from the sidelines now, but it's a case of judging based on the outcome, and not on the information available at the time (Taleb's canary calls notwithstanding).
The underestimation of the likelihood of everything going to hell at once is easy to criticize in hindsight, but very difficult to price for at the time. The fact is, a 20% default rate is ASTRONOMICAL, and out of all sorts of bounds of past history. Even a 10% default rate is pretty unheard of. Should those scenarios be accounted for? Yeah, in hindsight, it would seem so, but there is a razor thin margin on these deals, and as in transactions of any type (banking or otherwise), you don't expend resources on a possibility that is seen as extraordinarily remote, no matter how disastrous it may be. The incentives just aren't there for any one given deal.
That's why central oversight, to take these off of the OTC market and onto an exchange (which has been expected for months in the industry, and was officially announced by Geithner this week), will improve things right off the bat just by providing transparency and market incentives. How many caps and limitations are then necessary is up for debate, but I always prefer the carrot to the stick. Obviously some better measurement of how much risk is on the books needs to be done, but that risk needs to be limited not as a hard limit, rather within context. Removing all flexibility is too drastic. People have a much better handle on them than you're giving them credit for, and the failure was to see the big picture of the macro economy, which is reliant on someone having information that just wasn't available to any one banker or investor.
Swiss Mercenary on 27/3/2009 at 15:46
Quote Posted by Rug Burn Junky
That's just it, it was never really reliable. You can cherry-pick results to make him look good or bad (and with everything across the board down now, it's pretty easy to make him look bad), but it was never that simple, and it's a very lazy criticism. Anytime you predict the future, you're going to get some calls wrong.Sure, they wouldn't have gotten bonuses if the company had been allowed to go under, but after the bailouts took place, many of them would have left much much sooner if they hadn't been promised the money. What happened in many instances was "Look, if you stay on, we promise to pay you $X," they continued to work based on that promise, and then when time came to pay, everyone got up in arms, and said "Fuck you, we ain't giving you jack shit." Which would have been fine if it were said 6 months ago, but once you get people to perform their end of the bargain, it's pretty shitty to turn around and yank the rug from under them, and that's the inequitable part that I have a problem with.That was really where you were running into problems. That said, and without checking your math explicitly, you're on the right track. It's still a little bit of an overreaction, but I agree in principle.
Unfortunately, people working on the ground floor of one of the banks here were treated exactly like that. "Perform well, and you'll all get bonuses for Christmas" turned into "Bonus? What Bonus?" sometime in November.
Despite them actually performing well.
Or so a friend of mine who works there said.
Starrfall on 28/3/2009 at 15:22
RBJ Swisher is better than Nady.... your thoughts.
sh0ck3r on 28/3/2009 at 19:27
We are NOT your monkeys. Grow up.
