Skip

Its his fault. All his fault
February 24, 2009 8:02 AM   Subscribe

Well they finally found it. The one person to blame for the entire financial crisis. The man who made it easy to create and rate CDOs. David X. Li - The inventor of the Gaussian Copula
posted by JPD (93 comments total) 15 users marked this as a favorite

 
More like they found the guy who created an algorithm which was used long after problems with it were noticed and objections were raised about its continued use amirite?
posted by Science! at 8:06 AM on February 24, 2009


Yes. I was being a bit sarcastic. The real story is he created a tool used in models that was deeply flawed. Everyone knew about those flaws but no one was interested in dealing with them as long as the good times rolled.
posted by JPD at 8:11 AM on February 24, 2009


Personally, I blame the members of the eating/finals club that said "don't bother with that math stuff, an alum will get you a job on the street no problem."
posted by allen.spaulding at 8:13 AM on February 24, 2009 [4 favorites]


Whatever it is, a Copula sounds like my kind of mathematical concept. No wonder the financial system got screwed using it though.
posted by ghost of a past number at 8:16 AM on February 24, 2009


I just want to say that Gaussian Copula sounds like a character from The Cyberiad, and would also make a great name for a band.
posted by jquinby at 8:17 AM on February 24, 2009 [1 favorite]


The first problem was that the formula, although elegant, was based on faulty logic; it didn't take into account that financial data and relationships can change quickly and drastically. But, it spit out a number - that's all that management boneheads and beancounters these days understand or care about.

The bigger problem was they simply used this number as a way to devise bets and parlays; they became simply glorified oddsmakers. The market is a way to finance companies by allowing investors to take a stake, not a casino. Unfortunately, since they were stupid, lazy, and greedy to begin with, they wound up on the wrong side of the bet and now we have to bail them out.

Fuck them all. The system is rotted from the inside.
posted by Benny Andajetz at 8:17 AM on February 24, 2009 [1 favorite]


As Li himself said of his own model: "The most dangerous part is when people believe everything coming out of it."

Sigh. Are we hardwired to look for the perfect solution, the answer to all of our problems, something foolproof?
posted by anniecat at 8:18 AM on February 24, 2009


As long as he didn't plagiarizer it.
posted by bondcliff at 8:18 AM on February 24, 2009


plagiarizer? Can we have that three minute edit now?
posted by bondcliff at 8:18 AM on February 24, 2009


just want to say that Gaussian Copula sounds like a character from The Cyberiad, and would also make a great name for a band.

It also sounds like a great name for a car.
posted by anniecat at 8:19 AM on February 24, 2009 [1 favorite]


Pardon the derail, but I have to rant.

What's really devastating the economy is not the CDOs alone but also the credit default swaps that amount to bets that the CDOs will fail. Every time a CDO fails, a bunch of CDSs come due that are collectively worth many times the value of the CDO. Buyers of credit default swaps don't even have to own what the instruments insure. It's like someone else betting that your horse will break a leg, and then your taxes have to bail out the guy who bet that your horse wouldn't break a leg. It's insane.

So what's happening now is that bailouts of institutions that wrote CDSs become payments to hedge funds that bought credit default swaps. It's a transfer of taxpayer money to hedge funds. (ornate insect already linked to that in another thread but it bears repeating, pardon the pun)

The only sensible intervention is two pronged - one, governments have to nationalize the CDSs, maybe through some exercise of eminent domain - pay the investors pennies on the dollar of the original purchase price and then choose not to exercise their right to cash them in. Number two, and this is actually not as important, bail out the home owners to stop the CDOs from going bust.

The alternative is to bail out the financial firms that wrote CDSs they could never possibly cover to the tune of possibly hundreds of trillions of dollars.
posted by fleetmouse at 8:19 AM on February 24, 2009 [3 favorites]


Perhaps the people that used algorithms blindly weren't aware that in finance, past performance is not an indication of future results. From the article, it's not very convincing that the equation doesn't determine a useful score:

And because the copula function used CDS prices to calculate correlation, it was forced to confine itself to looking at the period of time when those credit default swaps had been in existence: less than a decade, a period when house prices soared. Naturally, default correlations were very low in those years. But when the mortgage boom ended abruptly and home values started falling across the country, correlations soared.

So the problem might have been a lack of data. That's the same way with most economic theories: there's no way to do experiments and get good data.
posted by demiurge at 8:21 AM on February 24, 2009


Voodoo economics.
posted by Faint of Butt at 8:22 AM on February 24, 2009


Gaussian blurry
posted by porn in the woods at 8:22 AM on February 24, 2009


Does this mean I can't blame poor black people anymore? Regards, Sean Hannity.
posted by billysumday at 8:22 AM on February 24, 2009 [2 favorites]


Sometimes hitting bottom is a good thing.
Obama wouldn't be president if it weren't for Bush.
posted by weapons-grade pandemonium at 8:24 AM on February 24, 2009 [1 favorite]


I was late for work because I started reading this article in the bathroom and couldn't quite get in the shower. (I actually get the magazine, yet didn't think about the portability of old media.)

Recommended to those even who don't "get" math as fairly interesting.
posted by MCMikeNamara at 8:25 AM on February 24, 2009


A. Sklar (1959), "Fonctions de répartition à n dimensions et leurs marges"

He's the ringleader. Came up with an existence theorem for copulas and everything.

But more importantly, I take issue with the idea of calling Li the 'inventor' of the Gaussian Copula. it looks like he derived the Gaussian from Sklar's original work - the Gaussian is the copula that works when the bivariate distribution is normal. He knew that it had to exist already, and did the work of finding out how to get to it. He's the discoverer of this formulation.1

And I think there's a big difference. Those who discovered atomic energy are not as blameworthy as those who used the discovery to invent the atomic bomb, who themselves are not as blameworthy as those who built & used it.

1. After looking through the pdf it seems that he didn't even come up with the Gaussian, but rather took tools that are in use by actuarial & other fields and applied them to default rates. Maybe makes him a little worse than I described, but not much.
posted by Lemurrhea at 8:29 AM on February 24, 2009


That VillageVoice article is a debacle. Nearly all of the damage CDS's have wreaked on balance sheets has to do with the fact that they must be marked to market as the value of the underlying CDO declines. Very little cash has been paid out, and if you actually look at the cashflow characteristics of the underlying CDO's a good argument could be made that very little cash will end up being paid out.
posted by JPD at 8:29 AM on February 24, 2009 [1 favorite]


Perhaps the people that used algorithms blindly weren't aware that in finance, past performance is not an indication of future results.

Modern corporate culture is run by what I call "driving by looking in the rearview mirror". Quarter-over-quarter comparisons are all that count. That's why why you see all this stupid shit like massive layoffs because profits were lower than expected. Not a loss, mind you, or even a horrible quarter- just not up to the beancounters' snuff.

As long as corporations don't have to pay any kind of social cost when they treat workers as expendable, we will visit this kind of situation again and again.
posted by Benny Andajetz at 8:31 AM on February 24, 2009


He's a mathemetician. He built a model. Mathemeticians are like engineers, they look at something and they pull it apart to find its characteristics, quantify, describe, and form hypothesises about their characteristics. Then they test their hypothesis and prove that it works in a theoretical case with real data. They model the past, prove that the model works, then perform some level of regression analysis and apply the model forward. When he put his model forward, it worked. However, ask an engineer when their project is over... there's always something new to model, always some special case, always some new characteristic to describe.

The problem is, in this case, people (by people, I mean financial people) didn't care that their behavior changed in response to the information the model provided. Moreso, they didn't care - they were making money and hell be damned - they can point the blame backwards that the model was an industry standard and they weren't doing anything that nobody else wasn't (... wait... I think I said that right... you get my point though...) Its about accountability, the model removed accountability, morality and common sense.

I'd imagine there's a ton more that David X.Li modeled in regards to this, but if it wasn't provable, it wasn't going to make it into the system - moreso, even if it was proveable, if it would have meant a sizeable investment to update the formula without making more money, there was never going to be a change - afterall... since everyone was doing it there was no reason for an individual institution to be responsible...
posted by Nanukthedog at 8:32 AM on February 24, 2009 [1 favorite]


What's really devastating the economy is not the CDOs alone but also the credit default swaps that amount to bets that the CDOs will fail.

Actually, the original idea was to produce hedges that would cover losses in the event that whatever credit agreement they covered would fail. The idea is, you're basically purchasing insurance on the loan.

What people started doing was placing bets that the perceived likelihood of a default would increase slightly and thus the value of a CDS would go up. No one ever expect all of the defaults to actually come through, and the result was the people (like AIG) who actually underwrote the original policies also became insolvent.

Sometimes hitting bottom is a good thing.
Obama wouldn't be president if it weren't for Bush.


Sure, but Obama is already president. The republicans are banking on people having a short memory, and blaming Obama for the problems we end up happening. It could happen, on the other hand, it may be that people have a long memory, and end up reacting to the problems by tilting even farther from to the left.

That's why people like Rick Santelli and other elites are so desperate to get people's anger focused on eachother, rather then the people who caused the problems.
posted by delmoi at 8:34 AM on February 24, 2009


Whoa whoa whoa--this is a family weblog! Gaussian tater.
posted by DU at 8:35 AM on February 24, 2009 [2 favorites]


So all the economy needs is a thorough degaussing, right?
posted by Sys Rq at 8:38 AM on February 24, 2009 [6 favorites]


It's all a blur to me.
posted by Poolio at 8:38 AM on February 24, 2009


Gaussian models don't work for financial markets. Just ask the boys of Long Term Capital Managment. For a more accurate model of financial risk, you gotta go with a Cauchy distribution. Just ask Benoit Mandelbrot.
posted by Fuzzy Monster at 8:52 AM on February 24, 2009


So it's fair to say the market GOT COPULATED HEY-O THIS THING ON I'LL BE HERE ALL NIGHT.
posted by The Straightener at 8:55 AM on February 24, 2009 [2 favorites]


Except the Gaussian part of the model was the default probability of a population - not asset prices.
posted by JPD at 8:59 AM on February 24, 2009


He's not a jew? Shucks, that's one theory out of the window. Are you sure he's not even a little bit jewish?
posted by MuffinMan at 9:03 AM on February 24, 2009


Every time a CDO fails, a bunch of CDSs come due that are collectively worth many times the value of the CDO. Buyers of credit default swaps don't even have to own what the instruments insure. It's like someone else betting that your horse will break a leg, and then your taxes have to bail out the guy who bet that your horse wouldn't break a leg. It's insane.

I'm so sick of these glib, half-baked diagnoses from people who just now figured some stuff out. The settlement processes for the huge Fannie/Freddie and Lehman defaults worked out just fine, probably because most protection-sellers weren't AAA and had to maintain collateral with their broker. The problem is that entities like AIG engaged in ratings arbitrage--they could write protection via CDS without having to post collateral and without being subject to traditional insurance regulations because they had a AAA-rating. While things were good, AIG could sell CDS and earn a payment stream without putting up any capital at all--that's the most profitable business in the world until it blows up.

The problem with CDS isn't that there is no ceiling on the notional value of the contracts, it's that there's no regulations in place to maintain prudent capital requirements. A centralized exchange that incorporated margin requirements would alleviate this issue, while still allowing for the benefits of the CDS market (namely, the ability to hedge/speculate on credit risk distinctly from interest rate risk and the resulting public price discovery).
posted by mullacc at 9:07 AM on February 24, 2009 [3 favorites]


I don't know who this David Li is, but I can authoritatively state that the financial crisis is not his fault. Because it's mine.

In November of 2000, I bought some mutual funds, and 2 days later, there was the election that never ended and a stock market "correction".

In October of 2007, I bought a condo. Presto, housing market crash.

In November of 2007, I sold some American dollar investments out of fear that the Canadian dollar would keep rising and they'd be worth less when I finally needed the money to pay for the condo I'd just purchased. That week was basically the all time peak of Canadian dollar value against US.

Last Friday, I bought some more mutual funds. This Monday, the market is the lowest in 10 years.

Every major financial decision I make is almost immediately follow by whatever market conditions are necessary to make my decision stupid. Usually this is a major crash and burn.

All the money and research that goes into analyzing this stuff is pointless. All you need to do is whatever I do. But a few weeks later, after I've crashed the market in whatever it was.
posted by jacquilynne at 9:08 AM on February 24, 2009 [8 favorites]


That's a great article. I am pretty sure this model, although probably not Li, showed up in one of the prior threads here. It's amazing how all these so called financial geniuses failed to really understand what kind of system they were building. It might even be worse in that they knew but didn't care as long as they continued to make huge sums of money.

The plunge.
posted by caddis at 9:12 AM on February 24, 2009


I'd like to point out that this is the third fpp today about blaming smart people for something bad. The quiz show woman is so smart that she's smug and makes everyone feel bad. The Iran post raises then dismisses rising female literacy as a possible cause of Iran's slowing birth rate, and now this Wired article blames a mathematician for coming up with a formula that works just fine as long as not a lot of borrowers all defaulted at once (which of course, they did).

This formula (and its use) is an important part of the backdrop or the context of this. It's interest is historical only. But so is the fact that everyone got greedy and sloppy - borrowers and lenders alike. The borrower who loses his house because he can't paid the absurd mortgage he took out deserves it just as much as the Lehman equities trader.

You can not make value distinctions between them because that does not help solve the problem. Doing it makes you feel good, but the problem is not that you feel bad. Now of this helps us right now.

I look at the situation as it has unfolded since August and I'm convinced we are going to fall all the way to the bottom of this without a safety net and without a parachute. People are approaching this problem morally, politically, sociologically, etc. Bankers are bad and ruined the economy. Homeowners are good and need to be saved, etc. Moralizing and demonizing everywhere, and still no evidence that the people in charge actually (a) understand the actual problem as it exists right now (as distinguished from learning the history of the problem - not the same thing) and (b) how they can fix it without sowing the seeds of another massive problem.

The government cannot stimulate the economy forever, and it can't offer mortgage relief to borrowers forever. This is the economic equivalent of intelligent design. At some point the system has to grow and evolve to produce the jobs on its own so that people can pay back their loans on their own, while simultaneously spending on their own the money that keeps the system alive.

Likewise, making villians out of Wall Street doesn't really help either. More than likely, the people who are capable of correcting this problem are the same ones who had a hand it creating it. We might have to learn to live with that.
posted by Pastabagel at 9:12 AM on February 24, 2009 [5 favorites]


From the Wired article:

The CDS and CDO markets grew together, feeding on each other. At the end of 2001, there was $920 billion in credit default swaps outstanding. By the end of 2007, that number had skyrocketed to more than $62 trillion. The CDO market, which stood at $275 billion in 2000, grew to $4.7 trillion by 2006.

Where did all that money come from? Anyone care to explain? I think this might be just as interesting.
posted by w.fugawe at 9:16 AM on February 24, 2009


Where did all that money come from?

Nowhere. That money doesn't exist. That's the problem.
posted by Sys Rq at 9:23 AM on February 24, 2009


Where did all that money come from? Anyone care to explain? I think this might be just as interesting.

Think about it this way. Let's say you buy $1 million dollars worth of life insurance for $100 a month. Then, the insurance company discovers that another insurance company would sell life insurance on you for $80 a month. So they buy that policy and pocket $20 a month free and clear.

Now the "market" on your life is a whole $2 million dollars, but you've only paid your monthly premium. The money hasn't 'come' from anywhere and it hasn't gone anywhere.

I'm assuming that's what's going on when they talk about the "$62 trillion dollar CDS market", which according to Wikipedia's article on the global economy is actually more then the GDP of the entire planet (measured in cash rather then purchasing power parity)
posted by delmoi at 9:25 AM on February 24, 2009 [1 favorite]


I'm so sick of these glib, half-baked diagnoses from people who just now figured some stuff out.

I'm sick of financial "sophisticates" who just destroyed the economy patting me on the head and telling me I don't understand, and we should just let them fix things up right nice. Oh, and when it's all over, they'll still be rich and I'll still have no health care, no job security, and no prospects for retirement. 'Cause that's how the market works. All hail the market!

Likewise, making villians out of Wall Street doesn't really help either. More than likely, the people who are capable of correcting this problem are the same ones who had a hand it creating it. We might have to learn to live with that.

On the other hand, there are 350 million people in the United States, and six and a half billion people on the planet. Why don't we put the couple of thousands people who created the problem out on the street and look for some folks out of those 350 million who can actually run a bank.

It's simple, guys. The people who said "we know what they were doing" were lying. That doesn't mean it's the fault of smart people everywhere, it just means that the "smart people" who were running our economy turned out to be liars, con men, and thieves. It doesn't have to be like that.
posted by vibrotronica at 9:36 AM on February 24, 2009 [9 favorites]


it's been mentioned a few times! cf. hsu :P
posted by kliuless at 9:45 AM on February 24, 2009


Likewise, making villians out of Wall Street doesn't really help either.

I have to confess that it helps me a little bit because, for a year or so back in college, I thought my heart was set on becoming an investment banker, when really, I had and still have no idea. But this makes me a little bit glad I never pursued it with the vigor and determination necessary to get into that field. Phew.
posted by anniecat at 9:47 AM on February 24, 2009


Kliuless - I wish I had seen your link before. That Hsu piece is great

I love this little factoid:

Investment banks, in order to figure out the rates of return at which to offer each slice of the pool, first had to estimate the likelihood that all the companies in it would go bust at once. Their fates might be tightly intertwined. For instance, if the companies were all in closely related industries, such as auto-parts suppliers, they might fall like dominoes after a catastrophic event. In that case, the riskiest slice of the pool wouldn't offer a return much different from the conservative slices, since anything that would sink two or three companies would probably sink many of them. Such a pool would have a "high default correlation."

But if a pool had a low default correlation -- a low chance of all its companies stumbling at once -- then the price gap between the riskiest slice and the less-risky slices would be wide.

This is where Mr. Li made his crucial contribution. In 1997, nobody knew how to calculate default correlations with any precision. Mr. Li's solution drew inspiration from a concept in actuarial science known as the "broken heart": People tend to die faster after the death of a beloved spouse. Some of his colleagues from academia were working on a way to predict this death correlation, something quite useful to companies that sell life insurance and joint annuities.

"Suddenly I thought that the problem I was trying to solve was exactly like the problem these guys were trying to solve," says Mr. Li. "Default is like the death of a company, so we should model this the same way we model human life."

posted by JPD at 9:56 AM on February 24, 2009 [2 favorites]


I, for one, blame math.

Every time a CDO fails, a bunch of CDSs come due that are collectively worth many times the value of the CDO. Buyers of credit default swaps don't even have to own what the instruments insure. It's like someone else betting that your horse will break a leg, and then your taxes have to bail out the guy who bet that your horse wouldn't break a leg. It's insane.

There were also so-called "synthetic CDOs," which were a secondary class of speculative securities derivatives (the guy on Diane Rehm today likened them to fantasy football). Basically, they were like a whole additional layer of pseudo-debt securities instruments whose performance was tied to the original debt securities they represented. These devices were also hedged by, get this: synthetic CDSs.

...

I'm so sick of these glib, half-baked diagnoses from people who just now figured some stuff out.... The problem is that entities like AIG engaged in ratings arbitrage--they could write protection via CDS without having to post collateral and without being subject to traditional insurance regulations because they had a AAA-rating.

Who cares about the details? What matters is this: Where is all the taxpayer money that's being spent on bailing these institutions out ultimately going? Are all the crazy synthetic default swaps being paid out to hedge funds and private investors? And if so, might it not be relevant if it turned out that the ultimate beneficiaries, especially of the more speculative derivatives, turned out to be private investors who also played decision making roles in institutions responsible for the irresponsible lending? Or for that matter, if they had direct or indirect ties to failed real estate investment firms with unusually high foreclosure rates?

I've seen no concrete evidence of chicanery, but then, I haven't seen any kind of evidence, because nobody's talking in any specific terms about where all the bailout money ultimately goes. But wouldn't it be one hell of a boondoggle if closer investigation turned up evidence of massive fraud in the unregulated securities markets? Without proper regulatory oversight in the first place, and without thorough investigations that yield a clear accounting of just who benefited from what deals now--in other words, without a clear picture of who ultimately ended up being on the receiving end of the payouts from AIG and others' unregulated swap and securities businesses, etc.--the general public is going to be out for blood for a long, long time.
posted by saulgoodman at 10:01 AM on February 24, 2009 [1 favorite]


From Wired:
Li's approach made no allowance for unpredictability: It assumed that correlation was a constant rather than something mercurial.

So this is the financial equivalent to the Tacoma Narrows Bridge? It works great as long as the wind doesn't blow?
posted by octothorpe at 10:02 AM on February 24, 2009


I'm sick of financial "sophisticates" who just destroyed the economy patting me on the head and telling me I don't understand, and we should just let them fix things up right nice.

Make an argument that shows you understand and maybe someone will let you try to fix it. I disagreed with what fleetmouse said, so I challenged it. If you want to argue the issue that's fine, but you don't know my background well enough to brand me with that scarlett letter.
posted by mullacc at 10:04 AM on February 24, 2009 [1 favorite]


Yay Waterloo. This guy and I were probably at school together.
posted by GuyZero at 10:06 AM on February 24, 2009


...an MBA from Laval University in Quebec. That was followed by two more degrees: a master's in actuarial science and a PhD in statistics, both from Ontario's University of Waterloo.
Blame Canada jokes in 3, 2, 1...
posted by howling fantods at 10:08 AM on February 24, 2009


Where is all the taxpayer money that's being spent on bailing these institutions out ultimately going?...And if so, might it not be relevant if it turned out that the ultimate beneficiaries, especially of the more speculative derivatives, turned out to be private investors who also played decision making roles in institutions responsible for the irresponsible lending?

I absolutely think these are worthwhile questions. People who bought protection from AIG were taking counter-party risk whether they knew it or not and there's no reason that taxpayers should be obliged to rescue those parties. We've done it for the sake of the system as a whole, and that's a decision open to challenge.

The issue I have is with demonizing the concept of a credit default swap. This may be a weak analogy, but I think of it a little bit like blaming the gun rather the the shooter. CDS may enable risky behavior, but they're only an instrument for those who sought to take on risk. If we try to solve the problem by eradicating the instrument, rather than intelligently regulating it, we risk losing all the benefits and doing little to solve the underlying problem (propensity to take on too much risk).

Without proper regulatory oversight in the first place, and without thorough investigations that yield a clear accounting of just who benefited from what deals now--in other words, without a clear picture of who ultimately ended up being on the receiving end of the payouts from AIG and others' unregulated swap and securities businesses, etc.--the general public is going to be out for blood for a long, long time.

Right on again. I just hope we have the wisdom to avoid cutting off our nose to spite our face.
posted by mullacc at 10:17 AM on February 24, 2009 [1 favorite]


They should have used the Francis Ford Copula.

sorry
posted by dirigibleman at 10:18 AM on February 24, 2009


Make an argument that shows you understand and maybe someone will let you try to fix it.

i didn't say I knew how to fix it. I said I'm sick of being condescended to by people who are arguing for a failed status quo that benefits them and harms me.
posted by vibrotronica at 10:29 AM on February 24, 2009


What - so now Alan Greenspan can die guilt-free?
posted by ericcmack at 10:40 AM on February 24, 2009


There were also so-called "synthetic CDOs," which were a secondary class of speculative securities derivatives (the guy on Diane Rehm today likened them to fantasy football). Basically, they were like a whole additional layer of pseudo-debt securities instruments whose performance was tied to the original debt securities they represented. These devices were also hedged by, get this: synthetic CDSs.

Yo dawg.
posted by ryoshu at 11:20 AM on February 24, 2009


If we try to solve the problem by eradicating the instrument, rather than intelligently regulating it, we risk losing all the benefits and doing little to solve the underlying problem (propensity to take on too much risk).

I think we're basically in agreement, mullacc, although you might find my own personal regulatory prescriptions a little drastic. Stiff regulations designed to prevent the markets from effectively incentivizing risky behavior are what's called for, IMO, because I think that's what the combination of all these various financial hedges and the relatively high short-term interest yield rates of the riskier investment classes has done: it's created, on paper at least, a system where you get to have your cake and eat it, too.

Riskier investment strategies pay more in the short-term because riskier instruments typically offer higher short-term interest yields. And once those investments are hedged to the hilt, they start to look like they aren't even all that risky in the long-term. So in effect, unlike in the real world, where excessive risk-taking is selected against as a natural consequence of a greater likelihood of failure, in the modern financial sector fantasy land, wild risk taking is, perversely, almost regarded as a kind of special talent.

So my radical solution: Don't let issuers offer higher interest rate yields on their riskier securities offerings at all. In fact, adopt stiff regulations to systematically discourage risk-taking across all investment markets. The purpose of swaps and other hedges is to allow investors to manage risk. Why should riskier investments offer greater short-term rewards in the first place? Sure, banks need capital, investors want to see big returns, etc. Who doesn't? But as much as we may all want such things, even the all-powerful free market can't make the possibility of "riskless risk-taking" logically or mathematically compatible with reality, even if there's no shortage of people out there willing to sell the impossible and twice as many more willing to buy it.
posted by saulgoodman at 11:21 AM on February 24, 2009


The government cannot stimulate the economy forever, and it can't offer mortgage relief to borrowers forever. This is the economic equivalent of intelligent design. At some point the system has to grow and evolve to produce the jobs on its own so that people can pay back their loans on their own, while simultaneously spending on their own the money that keeps the system alive.

You can't have an equivalent of "intelligent design" without first having an equivalent of natural selection, which is to say an actual theory that's proven and well tested and accepted by every single expert. I certainly don't see you offering one and pointing out how it's accepted by virtually every economist. Then again, I don't see anyone saying that the government should try to stimulate the economy "forever" either.

The problem is that entities like AIG engaged in ratings arbitrage--they could write protection via CDS without having to post collateral and without being subject to traditional insurance regulations because they had a AAA-rating. While things were good, AIG could sell CDS and earn a payment stream without putting up any capital at all--that's the most profitable business in the world until it blows up.

Wasn't AIG's AAA rating based in part on it's stock price, such that if it ever needed capital it could simply sell new shares to raise it?

I've heard (but don't know) that one of AIG's biggest counter parties was actually Goldman Sachs, and that the AIG bailout was basically a proxy bailout for Hank Paulson's former company.
posted by delmoi at 11:34 AM on February 24, 2009


That is a terrible terrible idea. You have to have risk. You can't mandate taking less risk.

Riskier investments by definition must offer higher returns or no one would be willing to make them riskier investments move society forward. It was risky investments that gave us electricity.


No the problem here was a temporal mismatch between when the rewards came and when risks came. That's what needs to be corrected. All of these structured products moved these timelines apart. A guy got paid out the year that he closed a deal involving cashflows that would be paid out over 30 years - with no downside other then being fired if his #'s were wrong (which there is no stigma attached to in Finance BTW) - but who cares if you get fired from a job paying 150k base if you made 1.5 mil in the process.
posted by JPD at 11:37 AM on February 24, 2009


Likewise, making villians out of Wall Street doesn't really help either. More than likely, the people who are capable of correcting this problem are the same ones who had a hand it creating it. We might have to learn to live with that.

But is it a wise investment to trust them to? That assessment requires far more sunlight than we have on the issue.
posted by butterstick at 11:44 AM on February 24, 2009


Wasn't AIG's AAA rating based in part on it's stock price, such that if it ever needed capital it could simply sell new shares to raise it?

"Access to the capital markets" is a key factor in the rating agencies' assessment of AIG and every other borrower. This highlights the extremely flawed nature of the rating agencies' methodology--sure, AIG could access the markets when it wasn't distressed and had a AAA-rating. But when it ran into trouble, it risked losing its AAA-rating and the capital markets shut it out. The circularity of this is obvious and rating agencies have long been considered a joke on and off Wall Street.
posted by mullacc at 11:46 AM on February 24, 2009 [1 favorite]


Riskier investments by definition must offer higher returns or no one would be willing to make them riskier investments move society forward. It was risky investments that gave us electricity.

That's the point. You're supposed to be less willing to do risky things when you're playing around with other people's money (as banks and institutional investors do).

Speculative investments in risky securitized debt instruments and unregulated credit default swaps didn't give us electricity. We need to start distinguishing between actual investment and the crazy self-referential games the financial sector plays with money.
posted by saulgoodman at 12:14 PM on February 24, 2009


no you aren't - you are supposed to be taking as much risk as the capital holders have told you to.

You are totally missing the point BTW - The issue with the CDO's and CDS's isn't that they were high risk - its that people were convinced they were very very low risk - like a government security - and they ended up being a little bit riskier.
posted by JPD at 12:20 PM on February 24, 2009


the capital holders

If I have my money in a bank, I get to say exactly jack squat about what I consider to be an acceptable level of investment risk for the money in my account. Technically, the bank is the "capital holder," I guess. But I don't want banks making risky investments with my money, so I'm for regulations that discourage them from doing that.

AIG, too. They're supposed to protect me from risk--not expose themselves to so much risk that they can't fulfill their core function. Again, no one gives them money and says, "here--invest this according to this risk profile." They manage that capital themselves, as if it were their own, when it's not. So stiff regulations to discourage risk taking investments of any kind seem pretty appropriate to me. (I understand these companies also engaged in a lot of off-book business through subsidiaries that led to huge losses; again, not the kinds of thing entities charged with the responsible stewardship of other people's money should be doing.)
posted by saulgoodman at 12:33 PM on February 24, 2009


Your money in a bank is insured - by the government

The AIG entities that sell life insurance were regulated by the government and are completely solvent.

The people whos money got pissed away either explicitly invested in products they knew took that risk or implicitly by owning the shares of companies that took that risk. If you lost money because of those two things then caveat emptor.

And again to repeat - this is a disaster because the products were supposed to be at the very safe end of the spectrum. Home mortgages have historcally been the abolute safest asset a bank can hold. All a CDO was supposed to be was a whole bunch of mortgages sliced and diced together. The math was wrong on the slicing and dicing is what happened.
posted by JPD at 12:44 PM on February 24, 2009 [1 favorite]


The alternative is to bail out the financial firms that wrote CDSs they could never possibly cover to the tune of possibly hundreds of trillions of dollars.

But..but...but... what about the stress test!!! We must Stress Test the Banks!! That's what the media line is. Shut up with your stupid makes-complete-sense logic and stress test!!
posted by odinsdream at 12:45 PM on February 24, 2009


“I don't know who this David Li is, but I can authoritatively state that the financial crisis is not his fault. Because it's mine.”

Hey, there is a major car crash every 13 minutes in the U.S. So we’ve got to get jacquilynne off the road.

“So this is the financial equivalent to the Tacoma Narrows Bridge? It works great as long as the wind doesn't blow?”

Aren’t bankers the guys who will lend you an umbrella as long as the sun is shining?
posted by Smedleyman at 12:45 PM on February 24, 2009


There are regulations that discourage them from making risky investment - look up how Tier 1 capital is calculated - it forced companies to put more capital (meaning a lower ROE) against riskier assets. The problem wasn't the lack of rules the problem was the rules were based on history and things changed.
posted by JPD at 12:45 PM on February 24, 2009


Well they finally found it. The one person to blame for the entire financial crisis.


LETS GET HIM!!!


*lights torch*
posted by MeatLightning at 12:56 PM on February 24, 2009


Aren’t bankers the guys who will lend you an umbrella as long as the sun is shining?

I think you've confused them with insurance brokers. But then, so did they.
posted by Sys Rq at 12:59 PM on February 24, 2009


All a CDO was supposed to be was a whole bunch of mortgages sliced and diced together. The math was wrong on the slicing and dicing is what happened.

That's not what a synthetic CDO is. Not to mention the structured investments AIG and others made off the books through subsidiaries, so that regulators had no clue how big their actual exposure was and considered them well-capitalized when they weren't. But this has gotten way off topic already, so I'll apologize and leave it at that. /derail
posted by saulgoodman at 1:00 PM on February 24, 2009


Not to sound like a dick - but you don't understand enough finance - a synthetic CDO is is also a bunch of sliced and diced mortgages. The CDS that make up a synthetic CDO reference a CDO that is a bunch of sliced and diced mortgages.

Google Put-Call Parity. Its kinda sorta the same thing but instead of using an option + a loan to create the economic equivalent of owning a stock - these use a CDS plus a loan to create the economic equivalent of owning a CDO.
posted by JPD at 1:08 PM on February 24, 2009


I don't have to be a dick or to understand "high finance" to know swaps aren't mortgages, no matter how you slice them. They're hedges against defaults on mortgages, so the fact that synthetic CDOs are made up of CDSs that "reference" sliced and diced mortgages doesn't make them = sliced and diced mortgages.
posted by saulgoodman at 1:26 PM on February 24, 2009


I don't have to be a dick or to understand "high finance" to know swaps aren't mortgages, no matter how you slice them. They're hedges against defaults on mortgages, so the fact that synthetic CDOs are made up of CDSs that "reference" sliced and diced mortgages doesn't make them = sliced and diced mortgages.

So? A CDO doesn't necessarily have to be sliced and diced mortgages. It doesn't make them automatically safe to back them with mortgages nor does it make them any less risky. All a CDO does is take debt and assign income servicing that debt to investors according to the risk that they're willing to take so they don't have to know the detailed intimacies of every asset market they want to get themselves into.

You really should consider taking the hints offered by mullacc and JPD. Yes they acted a bit like condescending jerks but you have to understand they have to deal with people repeatedly that wander in with the market stuff they've learned off Wikipedia (which isn't necessarily a bad thing, learning is good) and some article about the evil financial tool of the month from Slate and think they know it all (which is bad and they don't).

Don't get me wrong. I'm not saying don't contribute. I'm just saying don't go off all half-cocked. I like this stuff and I read about it a lot but I ask lots of questions and if I do take a stab at explaining something I try to leave myself open for someone to correct me (hopefully in a graceful way) rather than proclaiming that "it's like this". You'll find it's a much better way of going about things than being dismissed with your pat on the head.
posted by Talez at 2:10 PM on February 24, 2009


If the swaps pay based on the performance of a pool of mortgages then yes it is just exactly the same thing as pool of mortgages.

What you are saying is exactly the same thing as "Issueing fixed rate debt + buying flixed for floating swaps is different economically then issueing fixed rate debt"

A CDO = a pool of mortgages that pay at a certain rate. I pay someone for a CDO. The price/yield I receive above treasuries is a function the markets estimation of the probability of default. If I decide I want to eliminate the default risk I buy a CDS on the CDO from someone. I pay an ongoing premium to the CDS writer = to the premium I was receiving over treasuries for the instrument I bought. Effectively I now own a treasury. If my mortgages underlying the CDO default I get two checks in the mail every time a payment is due - one from the CDO, and one from the person who wrote the CDS. I have to send the CDS writer a check as well for my premium. If the CDO doesn't default I get a check from the CDO and send a check to the CDS writer. At the end of 20 years its over.

A Synthetic CDO = I buy a treasury from the US government + sell a CDS on an actual CDO to a third party (it might be the person who owns the orignal CDO, it might not - it doesn't actually matter to me. If they actually don't survive it just means my CDS lapses and my downside is I own a treasury). I get my coupon from the treasury + the premium from the CDO. If all goes accoding to plan I receive my two checks and get my principle from the government in 20 years. If the Mortgages in the CDO do not perform according to plan I have to make payments equal to missed interest payments over the 20 year time period + I have to pay back the principle as it amortizes.

It doesn't matter that its a swap - what the swap pays is still determined by the pool of mortgages.
posted by JPD at 2:13 PM on February 24, 2009


The architecture of the synthetic CDOs as laid out in the article linked by Saulgoodman is pretty reprehensible. SPVs (Special Purpose Vehicles)-- really?

Here's the meat, for the lazier followers of the thread:

an unknown bright spark within one of the investment banks came up with the idea of putting CDOs and CDSs together to create the synthetic CDO.

Here’s how it works: a bank will set up a shelf company in Cayman Islands or somewhere with $2 of capital and shareholders other than the bank itself. They are usually charities that could use a little cash, and when some nice banker in a suit shows up and offers them money to sign some documents, they do.

That allows the so-called special purpose vehicle (SPV) to have “deniability”, as in “it’s nothing to do with us” – an idea the banks would have picked up from the Godfather movies.

The bank then creates a CDS between itself and the SPV. Usually credit default swaps reference a single third party, but for the purpose of the synthetic CDOs, they reference at least 100 companies.

The CDS contracts between the SPV can be $US500 million to $US1 billion, or sometimes more. They have a variety of twists and turns, but it usually goes something like this: if seven of the 100 reference entities default, the SPV has to pay the bank a third of the money; if eight default, it’s two-thirds; and if nine default, the whole amount is repayable.

For this, the bank agrees to pay the SPV 1 or 2 per cent per annum of the contracted sum.

Finally the SPV is taken along to Moody’s, Standard and Poor’s and Fitch’s and the ratings agencies sprinkle AAA magic dust upon it, and transform it from a pumpkin into a splendid coach.

The bank’s sales people then hit the road to sell this SPV to investors. It’s presented as the bank’s product, and the sales staff pretend that the bank is fully behind it, but of course it’s actually a $2 Cayman Islands company with one or two unknowing charities as shareholders.

It offers a highly-rated, investment-grade, fixed-interest product paying a 1 or 2 per cent premium. Those investors who bother to read the fine print will see that they will lose some or all of their money if seven, eight or nine of a long list of apparently strong global corporations go broke. In 2004-2006 it seemed money for jam. The companies listed would never go broke – it was unthinkable.

O how I would gloat if there didn't seem to be a giant supply-side turd floating in my broke-ass consulting punchbowl as well.

BEGIN DIATRIBE
on preview, nevermind.

OK, I think it's out of my system. Please continue with all of the more educational discussion.
posted by snuffleupagus at 2:27 PM on February 24, 2009 [3 favorites]


Holy crap guys don't hate on saulgoodman, he's right. Thinking they're the same is exactly the kind of thinking that caused this mess in the first place. Let me introduce you to the dirty asshole bitch in low-liquidity/systemic-downturn markets: counterparty risk. By setting up a synthetic CDO with one entity, it is completely different than a straight CDO.

While Gaussian Copulas are flawed as an equation in practical terms, finding the valuation is exactly what the equation is supposed to resolve (but does poorly). The principle is that finding the covariance is tricky as heck, but it's very rarely exactly 0.0 or 1.0, and measuring that is incredibly complex.

Let's presume that the net risk for straight CDOs measured through standard-deviation magical financial black magick is x.

The risk for synthetic CDOs is: x + counterparty_risk.

They aren't the same. If you think it'd be the same because you'd get reinsurance from AIG so counterparty risk is negated, that's exactly the logic that got us to this mess.
posted by amuseDetachment at 2:29 PM on February 24, 2009


That's not what a synthetic CDO is. That's some crazy structured finance thing that no rational person should be buying.

I believe instruments like that might exist - but you'd have to pretty gullible to buy it. Once again you would deserve to be wiped out.
posted by JPD at 2:34 PM on February 24, 2009


What is the counterparty risk on a synthetic CDO? Your counterparties are the buyer of the CDS and the government..

Counterparty risk is a huge issue - but not in this particular case. If the CDS buyer disappears you lose out on the premium and you just have a treasury. Bad, but not the end of the world.
posted by JPD at 2:38 PM on February 24, 2009


JDP: AFAIK for CDSes, the premium is pre-paid, no? They don't pay out, they don't get insurance. (I mean, there is generally a lump-sum immediately and annual pre-payments)

Either way, though, for synthetic CDOs, the counterparty risk issue isn't the buyer of the security, the issue is with the writers. If the seller goes under, then the buyer of synthetic CDOs is out a ton of money (plus the risk of the underlyings as described earlier, of course).

The belief was, "hey, let's get reinsurance" to account for this, but the reinsurers simply didn't have the cash to pay out, so everything was a dirty lie.
posted by amuseDetachment at 2:46 PM on February 24, 2009


I think the way the payments work is variable by contract as are collateral posting requirements. Most of the vanilla single name stuff is pay as you go though.

But yes that was precisely my point re: Synthetic CDO's - yes if you are the guy who bought a CDS that was used to create a synthetic CDO and that entity doesn't have enough collateral then yes you could be screwed.

But if you are the guy who sold the CDS (i.e. own the Synthetic) you are getting what it said on the box - a CDO that will pay out based on a the slicing and dicing of a pile of mortgages.

The counterparty risk is held by the buyer of the protection. A Synthetic CDS is the seller of protection.
posted by JPD at 2:57 PM on February 24, 2009


As long as the treasury bond does not fail the synthetic CDO should mirror the real thing. The problem isn't differing risk, it is the multiplication of the pain when one of these things gives way. JPD's explanation is good; here is one just a bit more detailed.
posted by caddis at 3:01 PM on February 24, 2009


Gaussian Copula? Didn't he direct Apocalypse Now?
posted by turgid dahlia at 3:01 PM on February 24, 2009


... so CDOs and Synthetic CDOs are different then. I was just trying to make that one point.
posted by amuseDetachment at 3:03 PM on February 24, 2009


I'm confused (go figure). If the numbers are these:

U.S. Mortgages = $11 trillion
Total CDO Market = $4.7 trillion
Total CDS Market = $62+ trillion dollars

Why isn't this a Ponzi scheme, and who couldn't see it blowing up eventually?
posted by Benny Andajetz at 3:40 PM on February 24, 2009 [1 favorite]


This:
multiplication of the pain
was a helpful concept in getting a better grip on the above. (Thanks, caddis!)
I was thinking 'contagion' but 'multiplication' seems clearer.

Putting confusion over which instrument is which acronym, for those that fit this description:

That's some crazy structured finance thing that no rational person should be buying.

and to the extent that "structured finance" here means "intentional obfuscation" (ie to avoid regulation or artificially keep liabilities/losses off of a company's books)--why shouldn't there be some stronger response to this than (promised) increased transparency? Why permit the "SPV" arrangement above at all? I'm increasingly reminded of the Milken scandal. Isn't there any fiduciary duty along the way to the companies/people that these huge financial entities are taking money from--often through feeder funds and the like--and investing in CDOs/CDSes etc. here? Why can't we criminalize this kind of "structured finance?" Since it basically boils down to "lying to the investor."
posted by snuffleupagus at 3:47 PM on February 24, 2009


I'm increasingly reminded of the Milken scandal.

Three guesses on the firm that invented the CDO...
posted by ryoshu at 3:58 PM on February 24, 2009


"No the problem here was a temporal mismatch between when the rewards came and when risks came. That's what needs to be corrected. All of these structured products moved these timelines apart. A guy got paid out the year that he closed a deal involving cashflows that would be paid out over 30 years - with no downside other then being fired if his #'s were wrong (which there is no stigma attached to in Finance BTW) - but who cares if you get fired from a job paying 150k base if you made 1.5 mil in the process."

I used to work at a magazine where the ad men were paid their commission when the ad ran, not when the advertiser paid off their account. This led them to selling bigger and bigger ads to folks who were already behind on their payments, because, well, if you cut someone off until they pay, often you effectively end the buying relationship, and almost as frequently, they don't pay anyway.

Unfortunately, this also distorted the amount of risk that the ad men were willing to take on, and since the publisher needed these ad men to continue selling (at one point, giving equity in the business rather than payments), the magazine rapidly got into debt trouble. This happened a couple different times, and usually what would happen is that the folks would eventually pay up (though sometimes they were able to negotiate a pretty significant discount for paying in a lump sum), but because everything worked out (pretty much) no reform was made to the system.

After 9/11, there was a really huge collapse in ad money, enough that the magazine never really recovered fully.
posted by klangklangston at 4:21 PM on February 24, 2009


"Again, no one gives them money and says, "here--invest this according to this risk profile.""

Uh, actually, that's pretty much exactly what a lot of investors do. Part of the problem was that they had bad information about the risk profile. But it's one of those things that you know when you're deciding how much of your portfolio to put into stocks and what kind of stocks. I have a higher risk profile for my 401k because I'm younger and expect to ride out market variances prior to retirement.
posted by klangklangston at 4:22 PM on February 24, 2009


You are totally missing the point BTW - The issue with the CDO's and CDS's isn't that they were high risk - its that people were convinced they were very very low risk - like a government security - and they ended up being a little bit riskier.

And according to The Giant Pool of Money two things happened:

1) the government securities started paying shit yields

2) there was (a lot of) money to invest, so CDOs looked like a good alternative.

Why why why has not one other story on the genesis of the meltdown mentioned #2? It's really weird to me that there's focus on money lent but not where it ultimately came from.

Is it not important, not true or should I fetch my tinfoil hat?
posted by morganw at 4:50 PM on February 24, 2009


I'm confused (go figure). If the numbers are these:

U.S. Mortgages = $11 trillion
Total CDO Market = $4.7 trillion
Total CDS Market = $62+ trillion dollars

Why isn't this a Ponzi scheme, and who couldn't see it blowing up eventually?


Well the total amount of CDS outstanding is greater then the total amount of underlying credit its isn't to the scale your #'s indicated.

CDS can be written on any sort of credit instrument not just CDOs. And CDOs aren't just made up of home mortgages.

For example you can buy a CDS that insures against GE defaulting on its debt.
posted by JPD at 6:10 PM on February 24, 2009


Why why why has not one other story on the genesis of the meltdown mentioned #2? It's really weird to me that there's focus on money lent but not where it ultimately came from.

It was borrowed at historically low interest rates.
posted by Talez at 7:10 PM on February 24, 2009


And there's plenty of stories out there blaming Greenspan and low interest rates for asset bubbles like these.
posted by Talez at 7:12 PM on February 24, 2009


I used to work at a magazine where the ad men were paid their commission when the ad ran, not when the advertiser paid off their account. This led them to selling bigger and bigger ads to folks who were already behind on their payments, because, well, if you cut someone off until they pay, often you effectively end the buying relationship, and almost as frequently, they don't pay anyway.

For this hubris, think Lucent. Let's float huge loans to our customers which we know they can never pay back just so we can keep reporting record sales. Yay! oops..........
posted by caddis at 7:44 PM on February 24, 2009


I hope this David Li fellow burns in hell for what he did, right alongside that murderous Albert Einstein.
posted by mazola at 10:01 PM on February 24, 2009


Don't get me wrong. I'm not saying don't contribute. I'm just saying don't go off all half-cocked.

What the fuck are you and JPD even talking about? I made a very broad, half-serious comment about a general idea for a regulatory mechanism--a requirement that issuers of various kinds of riskier investment instruments not be allowed to return higher interest yields for riskier asset classes than for lower risk investments--and suddenly I'm spouting off making half-cocked claims about finance. Look back: I didn't make any spurious claims. I didn't even say anything about excess risk-taking being the underlying cause of the financial crisis (though there are people a lot more qualified than me who have made that argument). I only offered a modest (I even pointed out it was extreme) proposal for a regulatory approach that would provide a disincentive against excessive short-term risk taking.

And note, mullacc agreed with my earlier points.

Christ, what smug, self-important assholes people can become whenever talk turns to the mystical arts of finance!

Uh, actually, that's pretty much exactly what a lot of investors do. Part of the problem was that they had bad information about the risk profile.

Klang: Yeah, I do the same with my Roth IRA. I was talking about when ordinary banks, like Bank of America, invest their depositors' holdings. The point was that BOA, for example, doesn't ask me how it should invest my deposited savings. But it has a responsibility to make conservative, low-risk investments, as an FDIC entity.
posted by saulgoodman at 6:44 AM on February 25, 2009


And point taken about the role played by the ratings agencies' inaccurate risk profiles.
posted by saulgoodman at 7:14 AM on February 25, 2009


Why why why has not one other story on the genesis of the meltdown mentioned #2? It's really weird to me that there's focus on money lent but not where it ultimately came from.

the 'missing asset class' hypothesis has actually been floating around for awhile; its most recent proponent being ricardo caballero, cf.
posted by kliuless at 6:46 PM on February 25, 2009


oh and the best one paragraph description of the crisis :P

cheers!
posted by kliuless at 6:54 PM on February 25, 2009


« Older Sex, drugs and Islam   |   The Youngest Parents Newer »


This thread has been archived and is closed to new comments



Post