January 01
software engr, young at heart. coding from early age. "digital brain" but with lots of feelings too. writing here mainly to publicize a few key issues, let off some steam, & for the feedback. plz write me comments, very much appreciated!! even on old posts!! helps me gauge reader interest/ reaction & steer direction of new posts. oh, and IMs often make my day & I usually reply. and long IM conversations are my favorite.

Vzn's Links

My Links
vzn on politics
vzn on cyber/geek stuff
vzn on big luv
vzn misc essays
JANUARY 24, 2010 12:02PM

did broken math bring down the system? geeks, quants, etc

Rate: 4 Flag

hi all, Ive been wanting to write this post for a long time, ever since I found an excellent article in wired that was in fact a cover story. "the formula that brought down wall st" [3]. it talks about the "gaussian copula" function.

as a younger dude I had dreams of making it big as a "quant". (short for "quantitative analysis"). for many years Ive played around with the statistics of stock markets, and statistical analysis, and put in a level of work that might be a little embarrassingly large looking at it in retrospect.

its a very interesting area to work in, and its some  of the most unruly data you can possibly work with. I do think there are very slim trends that can be datamined out of it. but when you add up transaction costs [cost to buy and sell] and "slippage", the tendency of a stock to cost more than the published price, and then add up all the tax complexity, its a very difficult area.

there is a whole other line of thinking by Nassem Taleb, now a very high profile/fierce critic of mathematics [or the inherent limitations, and its deceptive risks], who argues that anything using correlation analysis in finance is mistaken. I tend to agree with this pt of view.

now, I agree with all these analyses, but during the crash/panic of 2008, a lot of different voices and ideas were heard about what "caused" our system to crash. it does seem clear there was some kind of systemic problem.

my bottom line/conclusion after a lot of reading and thinking on the subject. the basic issue seems to be vast mortgage pools made out of subprime loans that were a relatively new financial invention. therefore, our regulatory bodies did not keep up with the risks. clearly, a large subprime pool of stocks, after actively traded a lot, paradoxically/counterintuitively begins to take on the risks of an index of stocks.

so roughly, we had rules to regulate this that were either very weak or poorly thought out, not thorough, patchwork. it was a hodgepodge. there was no agency that you could point to in particular that was responsible for the failure.

in my opinion the Fed deserves the most blame, but again there, the administration/headed run by Bernanke [who at the moment, I am strongly hoping/advocating he not be reconfirmed] has resisted strongly the implication that it failed spectacularly. this is the familiar bureacratic excuse, "its not my job" at the level of entire government agencies. "its not my department". "its not my agency"

think about it. in a well designed system, it wouldnt matter if individual corporations used wrong math formulas to value their assets. isnt that the role of regulation? Too Big to Fail is a massive Fail itself. its the role of regulation to prevent the kind of systemic failure that we exprienced in 2008.

so yeah, math is powerful, amazing, and magical at times, and the quants increasingly run wall street, but you cant really blame quants for the problem of systemic blindness and "not my agency" (by the way, "not my agency" is also virtually exactly the same problem we have with the multibilliondollar so called "intelligence" complex, but thats a whole other can of worms....).

we need a regulatory regime that can deal with all the fast paced "financial innovation", some of which of legitimate, much of which is not. the regulatory system's job is to be dubious, recognize smoke and mirrors when it finds it.

it would also seem we need an expert in "bureacratic architecture". we already have multibilliondollar regulatory agencies who historically, effectively add up to something that is less than the sum of their parts!! how do you build a system that coordinates multiple agencies and avoids holes and cracks between them? this is a sort of deep zen question that nobody in government asks. because they are like fish swimming in water. the invisible ubiquitous water is "agency glasses" or "agency lens".

and the idea of creating a new consumer protection agency is probably a feeble and destined-to-fail idea, in my opinion. agencies already existed whose job it was to regulate, and they failed. you do not solve that problem by introducing a new agency. but thats washington, if you want to make a bold move, you suggest that we need a new agency....

The Minds Behind the Meltdown
How a swashbuckling breed of mathematicians and computer scientists nearly destroyed Wall Street

How maths killed Lehman Brothers
by Horatio Boedihardjo

Recipe for Disaster: The Formula That Killed Wall Street

Why nerds must rescue the American economy
Lack of math skills contributes to consumers getting ripped off

Darpa: U.S. Geek Shortage Is National Security Risk


Your tags:


Enter the amount, and click "Tip" to submit!
Recipient's email address:
Personal message (optional):

Your email address:


Type your comment below:
It would be nice if there were one, single, simple explanation.

My personal opinion is that it was a catastrophe because securitization worked too well for too long. That is, starting in roughly the 1980's, securitization of debt (primarily mortgages but after a while everything) worked well. The most dangerous ideas aren't wrong ideas. They are good ideas that become too popular, are pushed too far, and break only after they are widely spread and utilized.

In terms of prioritizing problems, I would rate them as follows:

1. Optimistic assumptions. What is the probability of residential real estate crashing? Way, way higher then anyone thought possible.

2. Reliance on historical data that excluded panics and depressions.

3. Belief in markets. A naive belief that modern financial markets are liquid and deep enough to be self correcting over a wide variety of scenarios. We found that in a panic, liquidity disappeared in an instant.

4. Accounting "innovations" that tried to implement derivative accounting across all businesses. The result of this, combined by #3 meant that accounting rules, rather than being conservative and stabilizing rather reinforced market cyclicality.

6. Over reliance in derivatives and unregulated derivative market prices. Extremely thin, unregulated markets drove all asset prices in a totally unintended manner.

7. Feedback loops between rating agencies, derivative prices, and asset values. It became possible to short a stock to zero in a matter of hours. It was also possible to speculate on corporate failure using enormous leverage. After failing miserably regarding ratings on complex debt instruments, the rating agencies "got religion" and systematically destroyed asset prices by newly aggressive rating policies.

This is my personal take on the subject.

As far as a cure? Fortunately, I think that no one will buy new financial innovation for the next 20 years or so. We will never see another CDO^2 in our lifetime.

Remember that the most regulated financial entity in the history of the universe was the GSE's. Fannie and Freddie had dedicated regulators, numbering in the hundreds, and regulating only those companies. They were directly supervised by Congress.

Based on this, the idea that we need a lot more of that sort of regulation isn't very persuasive.
Bad math, per se, is pretty low on the list. Yea, they should have used thicker tails. And yea, they should have known that in a panic, correlations go to 1.0. But this is way lower than the institutional and epistemological issues that I see as the fundamental drivers.
thx for the very informed comments. I would tend to agree with your list.
wall st is anti all regulation, however. in the same breath you say that CDOs are toxic and will never happen again for 20yrs, but that regulation is useless.. wouldnt good/effective regulation have uncovered the mutant CDO issue and raised alarm bells? CDOs were "off books" in many cases, massive exposure not found in SEC statements.. true? "shadow banking system".... the point of effective regulation is to prevent a "shadow banking system" (antithetical to financial *transparency* and thereby stability) from even arising in the 1st place... who was asleep at the wheel?
"thicker tails"--- is that nongaussian?? yeah taleb is really emphatic on that point which I agree... financial returns can be fitted convincingly, without a lot of error, to gaussian curves but they are inherently nongaussian.... as for correlation....yeah... during a crash that which was correlated becomes anticorrelated and vice versa.... upside down and downside up so to speak....
elliot spitzer has an NYT editorial where he faults POOR ENFORCEMENT OF PREDATORY LENDING PRACTICES.... fannie/freddie are at the center of this..... the regulation there became blind to the issue, not sure exactly how..... "liars loans"....
in academia there is a concept/name for regulatory malfunction due to greed... its called "regulatory capture"... basically the industries/corporations that the us govt agencies are intended to "keep in check" end up getting high influence, and almost buying favorable reviews.. this happened with the ratings agencies you mention to some degree also.... in my opinion regulatory capture is highly analogous to biological models where the parasite alters or even takes over the host function/operation....
one symptom/aspect of regulatory capture is people that tend to cross the lines of public vs private frequently in their careers, making the membrane more "permeable" (weak) so to speak... ppl like geithner, summers, bernanke all fit this profile perfectly... another example is rubin...
ok CA, you have a point and its a worthwhile analogy to explore. if math formulas fail so badly in finance, how could they work in AI? but I would say, its a straw man argument. math formulas do in fact work very well in finance within a limited domain [the argument is over what exactly the valid domain is, not whether there is a valid domain], and will never be thrown away, and will be used forever into the future in that area. what they fail at is accurately predicting the future, esp far into the future (eg years), and esp when trillions of dollars are at stake and a rounding error goes into the billions, as was the case with the crash of 08 in various cases....

as for AI, I do believe there is something akin to a "formula" for it, and once you begin to work with any complex system that has sophisticated emergent behavior, I think you will begin to appreciate the power of symbolic/mathematical tools/technology. and it is indeed a technology, that continues to evolve. todays math technology is insufficient to generate AI, but it is progressing quite nicely, and I have strong confidence that tomorrows technology will achieve it, although not necessarily in my lifetime. and to say that something will not happen in ones own lifetime, but may happen after it, is actually a fairly conservative statement that I think few people could disagree with, who are familiar with the march and occasional outbursts/breakthroughs of technology.... ps have you read any Kuhn yet?
for example... a sophisticated artificial system that might be somewhat close to AI from today is.... weather modelling. consider the very large simulations that run on supercomputers. they exhibit emergent behavior-- engaging in trends not predictable by the scientists that built them, in behaviors not apparent from looking at the equations. have you ever heard of the lorentz equation? its a wondrous formula that almost singlehandedly caused a scientific revolution. I studied it in my sophomore statistics class in 92 or so-- have you played with it? only a generation ago, a scientist even working in the weather prediction field might have told you that the accuracy that is now achieved daily would be impossible to achieve....
so, CA raises an interesting paradox and one that deserves further scrutiny. suppose that we did create an AI system in the future sometime. could it have behaved smarter than the quants who blew up our economy in 08? perhaps. but another possibility is that the AI could indeed exist based on a math formula, yet it would be no better at predicting the far-off financial future than humans.

so in other words, paradoxical as it sounds and as I admit [great truth is often paradoxical!!] my position is, there exists a wondrous/near magical formula which we havent discovered yet that is capable of generating AI, yet it is not superhuman or supernatural in the sense that it can break basic laws of science or physics, such as that it cant predict the future. some limitations are merely speed limits that can be broken in the future, and some limitations are truly insurmountable, such as the increase of energy in the universe, other thermodynamic laws etc. -- and I bring up thermodynamics on purpose, because to me some of the inherent uncertainty in predicting the future may be so basic as to be almost a law of physics or thermodynamics, related esp to rise of entropy in the universe over time....

Fannie and Freddie were not the leaders in the rush to the bottom, but rather reluctant followers.

I am not opposed to regulation, but rather am concerned that it simply won't work.

I am optimistic in the intermediate term, because I don't think it has to work. That is, there is no primary market for asset based securities of the type that proved toxic.

Once you accept that we have business cycles that haven't been tamed and can't be totally tamed, you then try to determine whether regulations and market interventions are pro or counter cyclical.

In 2005 and 2006, lenders should have been much more cautious. Right now, it appears that they are too pessimistic and their reluctance to lend is inhibiting the recovery.

People are angry at banks. However they need to decide if they are angry because they are too reckless NOW or too conservative NOW. If it is the later, then trying to develop regulations that would have worked 4 years ago will just prolong the recession.
With regard to the "shadow" banking system:

Lending can be split into loans that are originated to be held vs. loans that are originated to be sold. Vanilla securitization is not a problem. It is pretty effective and not particular complex. The more complex financial products -- like "inner CDO's" are extremely complex and have no real rationale for existence.

Unfortunately, the so called shadow banking system has grown to the extent that traditional banks simply can't grow fast enough to take up the slack.

All the shadow banking system really consists of is a deconstruction of the lending process, with the various pieces handled by non related entities. That is, loans were originated by brokers. They were then purchased by investment banks, packaged into asset based securities, and then sold to fixed income investors, including pension funds.

This is a very efficient system, since the overhead is lower than traditional lending institutions. On the other hand, the intermediaries have few incentives regarding loan performance.

It is not unlike those gyros chunks of lamb. In theory, you can grind up the meat and then stick it back together in a very efficient shape, as well as having a uniform consistency. If you lived in a small town and knew who was involved in the process, then it would probably work out fine. However, once you move to a large urban area, and one person buys the raw material, another grinds it up, and sells it to someone else that puts it back together, finally selling it to a guy that is selling it out of the back of a van, then you are asking for trouble. It could still work out fine, but.... For me, I would prefer to just buy a lamb chop.

Likewise, I think a lot of pension funds and buyers of fixed income securities would now prefer to simply own old fashioned bonds.

I don't have any great ideas for running the shadow banking system. We seem to be able to live with fairly long and fragmented supply chains for many manufactured products, so there is no inherent reason it can't work.
wow, @#%$& spammers already. luv it.
I wrote "some limitations are merely speed limits that can be broken in the future, and some limitations are truly insurmountable, such as the increase of energy in the universe, other thermodynamic laws etc."
astute readers would probably realize, this is a [serious] typo & should read
"some limitations are merely speed limits that can be broken in the future [eg lack of AI formula], and some limitations are truly insurmountable, such as the increase of *entropy* in the universe, other thermodynamic laws etc."
Wow, this explains a lot. I'm very confused by all this and wonder how you sort it all out with out feeling like your on some weird series of interconnected hamster wheels.
There is much I don't understand, like how fractional reserve banking which creates more money is good. How does me depositing 100 bucks so a bank can lend maybe 90 and get back 110, so they "create" and get to keep the money which is now put back into the economy. There it is "extra money" and what the bank produced was debt/or takeaway for someone else. The bank created money by creating debt?

How does a bank financial statement list someone who might pay them back the $90 on the asset side of the sheet with the $10 of mine they still owe? How is money you don't know you're going to get the asset? On the other side my 90 bucks they owe me is a liability and when they stop paying they now have my entire $100 is their liability? No matter what math I apply, the concept that this financial statement shows soundness seems foolhardy at best.

How is me sending my money to some 401K in WallStreetWorld going to benefit me if it's investing in funds that may invest in my company. Then my company sends my job to wherever and I'm losing money again while I'm job hunting again. But I do get a pittance in interest in my 401K because what makes corporations wealthy is sending jobs overseas. What makes investment bankers wealthy is the same thing. Me unemployed. Or me working more for less to satisfy my need for just 1% more in interest. How can I know which funds invest in companies that pay people like me to work in the U.S. at decent wages.

Do the experts believe we don't know the subprime loans not getting paid were not a surprise. What person in the right mind gives someone a loan where the payment will increase beyond what they KNOW the persons income will rise. Bull, they figured they'd get the houses back, keep the interest they earned as people (the filthy masses) struggled to keep paying. I can almost see the blood dripping from the fangs as they drain the life from those people. Doing my 4 years in Real Estate I went to a lot of closings. Stacks of papers to protect the buyer from the vampires, really were those people going to sit and read an inch thick stack of loan docs. The bankers gambled that the buyers wouldn't be able to re-fi, the market would go up and they'd re-sell the repos at a profit.

I know it's hard to believe they are that empty but look at those CEO's. Jamie Dimon was "a little tired" of the CEO bashing? Really, my single mother friend is a little tired of trying to enjoy foreclosure. Moynihan whimpered that the majority of the people getting bonuses didn't "cause" the crisis. Really, sorry must have been my fault. They will divvy up what, 145 Billion in Bonuses they DESERVE because they are talented? While this is my 13th month since I was laid off. I must not be talented or deserving. I feel a little like sticking them with a pin to see if they flinch.

145 Billion they all deserve, while Haiti is crushed. The all important Investor is having a collective tantrum. They want to gorge and gorge, while their money works so they don't have to, while Haiti is crushed. Why should I care if they lose all their money. They can just modify their home loans.

I don't understand any of what you have written. I'm just not that educated - 9th grade, and I'm not that smart. But I know that worker means human. I know that an unemployed worker is usually a non-consumer so the B.S. doesn't fool me.

I will not feel safe as an American until these banks all fold and take Wall Street with them. These banks are too big to succeed, just like the big bureaucracies. Too big to succeed, like the regulators. If we didnt' have these crafty freaks we wouldn't have to waste money on all these regulators. They cause the regulations, then whine about the taxes. They lust with passion over my tax rate, talk about envy. If they want my tax rate, let them work for 20K a year. Unfortunately the jobs I've had are hard, I don't know that they can take the abuse and the stress.

I keep my money in local banks to loan to local businesses. I don't care if I have a shit pay/bene job at Macy's or the local boutique. The only benefit I have is if it's local. I'm glad one of our two Macy's folded, then people will have to buy on Main Street. Screw Macy's, if they survive fine, if not there are plenty of crappy jobs full of managers who go to meetings with important executives to talk about getting me involved in thinking outside the box. They took my cheese and moved it out of the fridge and now it's too moldy for me. To Hell with all these big corporations.

I am frustrated, each year I fall lower in income and I've worked my tail off for 38 years to get here. I'm a disposable, I'm getting disposed of more often. I wish someone would explain which of the formula's will help me make any sense of this. At my last job I got really lucky, old people like me will kill themselves with work ethic. The young people there really helped me get past being loyal and working my tail off so it doesn't feel like betrayal when you get treated like nothing. Like the day I went into my supervisors office after being put on salary with no raise because we had a project with overtime (Exempt employee) Mandatory 8 to 8, m-s, 8-5 sun. Three months of that project and we rocked. I asked her for my overdue review for a raise. This is what she told me as she was typing...

Here at ____ we do things in the business world, like when I was a manager at Macy's (spiral wristband with a key chick). We pay the average person like you what we would pay any average person doing an average job like yours.

I will never forget the sound of her voice or those words. I don't want to. That was the moment those kids reached me. I'll never again give so much of myself for so little money to those who get so much money for doing so little. That is a formula burned into my brain from her words. The rest of it, I really tried to understand but it's over my head. Thanks anyway for asking. It made me feel like I mattered.
l'Heure Bleue:

I will assume that you are serious about wanting to understand some basics about banking.

First, some basics on bookkeeping. Assets=Liabilities+Capital.

For vanilla banks, liabilities are deposits. You put money in the bank. For you it is an asset -- for the bank it is a liability.

If you make a deposit of 100, the bank has an asset of cash=100 and a liability of your deposit=100.

If a bank has 10,000 accounts, and they all have average balances of 100, they have 1,000,000 in deposits. People put money in and take it out, but overall, the deposit base is relatively stable at 1 million.

Lets say the bank lends out 100,000 to a customer, using a house as collateral. The loan is a liability for the customer, but an asset for the bank. So the bank trades an asset of 100,000 in cash for a different asset -- a 100,000 loan paying interest.

How is a loan an asset to a bank? It is the same as a bond or the promise from a customer to be paid in full, with monthly interest as well as the full principle.

Now, if the customer quits making monthly payments on the loan, the bank will take and sell the collateral. When a loan is overdue, the bank estimates how much it will recover by selling the collateral. Any shortfall from the principle amount of the loan will be treated as a loan loss reserve, which is a "negative" asset.

The banks balance sheet is the mirror image of their customer's balance sheet. That is, if you are a customer of bank XYZ, your deposit is an asset to you and your loan is a liability to you. For XYZ, the loan is an asset and the deposit is their liability.

I seriously doubt if this will clear up your more serious questions regarding the economy. But the arithmetic of banking is pretty simple when it is explained.
As far as I know, Wired was the first to publish the story on the formula that brought down Wall St. Although it was interesting reading, knowing how we got screwed did not make it any less painful.
LHB.. as bill clinton said once, "I feel your pain"....
yeah, I share your distaste/distrust for the modern corporatocracy.... (just reaffirmed in a dangerous way by the supreme court re financing!!)
you write a bit of a near-conspiracy theory, but you make a very important case that many economists and politicians and the public gloss over in obliviousness to the basic dynamics of our financial system. there are some cases where banks can indeed profit on foreclosures if their margins are "good". this introduces a variation on what in finance is called "moral hazard" where, defined loosely/generally by me, economic incentives go in the opposite direction of moral, health, or quality of life considerations. there are many cases of this in capitalism, and economic theory is mostly deaf, dumb, and blind to this farreaching implication.
yes, corporations profit on a kind of arbitrage against humanity. dumping toxic sludge into the environment, or into the stock market!! .. etcetera... all increases their numerical bottom line. unless the govt keeps them in check effectively. again, moral hazard. the accounting must be off in this case. it is the govts job to line up the accounting so that increased revenue not synonymous with increased tyranny and spoilage. govt fell down on this role....
NC -- I largely reject some of your statements although we may agree on some basic principles. I agree its critical to regulate the right areas or its useless, and the govt may easily get this wrong, because its a matter of the utmost delicacy. its rare I would do a point-by-point rebuttal, but your msg/generalizations seem to evade reality....

"Fannie and Freddie were not the leaders in the rush to the bottom, but rather reluctant followers."
doesnt make sense to me. why did they go bankrupt? who were they following? if they are effectively underwritten by the govt, then they MUST follow the orders of the govt, and those orders must be inherently conservative ensure their safety EVEN in a downturn or recession. they need to act/behave & be regulated like BANKS and not venture capital organizations.

"I am not opposed to regulation, but rather am concerned that it simply won't work."
why wouldnt it work? I agree it has to be aimed carefully....

"I am optimistic in the intermediate term, because I don't think it has to work. That is, there is no primary market for asset based securities of the type that proved toxic."
this seems a dangerous concept, that if there is a market for [x] then the govt should not really regulate it. this is part of the flawed thinking that led to the crash and the failure of the securities. financiers are of the opinion, if I can sell it for [x] it is worth [x]. this is true for stocks-- but *not* commodities like beer, cereal, houses.... it is clear, that the value is determined by how well it holds up over time, an unanswerable question at the time of sale.

"Once you accept that we have business cycles that haven't been tamed and can't be totally tamed, you then try to determine whether regulations and market interventions are pro or counter cyclical."
regulation exists to prevent fraud, eg bernie madoff. and a so-called "bank" investing in risky investments that can all blow up if the GDP stops going up exponentially....

"In 2005 and 2006, lenders should have been much more cautious. Right now, it appears that they are too pessimistic and their reluctance to lend is inhibiting the recovery."
this lumps in all banking activity into either being too liberal or too conservative. ie cyclical. but banks, underwritten by the govt, have no valid business being involved in anything with cyclicality whatsoever.

"People are angry at banks. However they need to decide if they are angry because they are too reckless NOW or too conservative NOW. If it is the later, then trying to develop regulations that would have worked 4 years ago will just prolong the recession."
I say, its a false dichotomy. regulation recognizes the basic concept that just because a bank can buy [x] and a market for [x] exists, does not mean that is a valid purchase by the bank. also, it requires that such transactions once made must be put on a balance statement somewhere and reported to the SEC.....
lets look at the lack of regulation and how it led to catastrophe. I have not studied fannie/freddie situation that much, but have read quite a bit about AIG.
it would seem that we have some basic categories of financial instruments
- cash
- bonds
- insurance
- stock
- mortgages/pools
- derivatives

now what financial organizations want to do is break down/blur the barriers between all of these and make them all behave like each other. the claim is that this increases "liquidity" ie makes everything behave like cash--stable. but instead, it has the effect of making everything act like stocks--volatile. ie if you bundle up mortgages into large pools and buy/sell them, you no longer have the stability of simpler mortgage instruments. you get something that behaves more like stocks-- up and down, volatile.

this is not financial innovation.... its more like financial incest....
so you get frankenstein monsters/abominations like AIG. they wrote "insurance" for "mortgage pools" but barely had any capital to cover it. so this is wrong because an insurance organization needs sizeable capital stored from its insurance payments/premiums. and mortgage pools are more volatile than unbundled mortgages-- they behave more like stocks. basically, you get the worst of both worlds so to speak because of regulatory confusion.

the role of govt is to try to keep the lines from blurring, and it does this effectively through regulation.
CA--- "Maybe we will be able to zero in on some of these axiomatic differences as time goes on..."
its not really difference in axioms, its a difference in experience. we approach the same issue from completely different directions. I am trained as an engineer, you are trained as a [x]. I dont think we can resolve the difference. I dont think you are reasoning from axioms. I think you are reasoning from feelings so to speak. what psychologists call "motivated reasoning". your a priori conclusion is that AI is impossible, because you find it distasteful, and you work backwards from this conclusion. that is my conclusion. =)
I cannot prove this about you, but there is a lot of snippets of evidence for it in your writing... as for me I reason largely by analogy, which I agree is imperfect, and admittedly not really the same as reasoning by axioms.... anyway, our debate is unresolveable in the short term.... all we can do is revisit it periodically to see what is now on the table, & what has changed....
I am confronted with my ignorance.
:( I know so little. I have tried to read about the meltdown and have gleaned a few things. One is that they just recertified and recycled junk bonds and unstable stocks in order to get a higher rating for them and make them appear safe, which made even conservative investment firms by into them and suffer. But I have so many questions about other things I read.
the basic instability, as learned from the original Mother of All Crashes, the Great Depression is:
borrowing to invest in stocks.
this is also called "buying on a margin".
and it can be regulated, and is strictly regulated today. you are strictly limited by law in how much you are allowed to invest in the stock market on margin.
this is the fundamental regulation that govt must enforce. (it is also very similar to a "reserve ratio" applied to banks).
now, how does this basic rule apply to AIG? it would seem that they were selling "insurance" on "mortgage pools". but those mortgage pools were traded like stocks. and the insurance was really, "creating some capital reserves to pay off in case of a drop in the market". but in fact, this is all basically identical to---
borrowing to invest in the stock market. its a basic pattern. why didnt our regulatory system recognize this?
answer-- regulatory capture, and smoke and mirrors (subterfuge) masquerading as "financial innovation"..... but that is what effective regulation would penetrate.....

Most of the things you brought up are simply matters of opinion. I could marshall quite a bit of information to support my opinions, but it still boils down to opinion. However, I will suggest how one could bring additional information to bear on the issues:

"Fannie and Freddie were not the leaders in the rush to the bottom, but rather reluctant followers."
>doesnt make sense to me. why did they go bankrupt? who were they followingthis seems a dangerous concept, that if there is a market for [x] then the govt should not really regulate it.regulation exists to prevent fraud, eg bernie madoff. and a so-called "bank" investing in risky investments that can all blow up if the GDP stops going up exponentially....this lumps in all banking activity into either being too liberal or too conservative. ie cyclical. but banks, underwritten by the govt, have no valid business being involved in anything with cyclicality whatsoever.
My lengthy reply got chopped off.

Let me just leave it as it stands.....

Interesting discussion. Remarkably civil for the internet also.

Rated, by the way.
hey NC, you are knowledgeable and informed on the subj and thats not common. sorry you lost your reply. I encourage you to rebut me point by point if you have the time.
you say its a matter of opinion. somehow, I disgree. in 2008, it was a Near Death Experience for the mass economy. there can be little question that recessions are periodic or "cyclical" as the whitewashed euphemism goes, and not really avoidable, but the *severity* of the recessions must be, surely? and surely, this was a very harsh recession due to the dynamics of our political/economic system at the time.... ie anti regulation.... now this, its hard for me to regard as a mere matter of opinion.... its a matter of *legality*....
ps I just read that geithner [who I wish would resign!!!], at his congressional hearing, agrees that AIG had "tentacles" all over the place, and was poorly regulated, that it basically slipped through the cracks.... now, whose fault is that??? and why arent we fixing those cracks???