Noam Scheiber at The Stash explains why the AIG bailout isn't close to being over. He notes that the bailout money is quickly funneled through AIG to its investors thusly (the emphasis below is mine)...
the payments arise because AIG has to post collateral every time the bonds it insured take a hit. (Credit default swaps--the financial instrument at the source of most of the problems--are basically an insurance policy for bonds. So AIG sold insurance to, say, Goldman Sachs on some bonds Goldman bought. And, as the bonds lose value, AIG has to post collateral to Goldman per the contract they signed.)
That's scary, because as the economy continues to deteriorate, all those bonds AIG insured are going to keep deteriorating, too, and AIG will have to keep posting collateral. Which means the taxpayer, assuming we don't let AIG collapse, is going to have to keep forking over cash.
That's pretty awful news already. So how could it get worse?
Take a look at that section I bolded above. The part about credit default swaps being insurance policies for bonds. By reading that you would assume that someone had to own a bond before taking out such insurance. Schieber's own example runs with this assumption. Would that it were so straightforward and confined.
The problem is that - and you might want to be sitting down before reading this if you're one of the taxpayers ultimately going to be tapped to cover this stuff - AIG could and did sell credit default swaps to all comers, regardless of whether they actually owned the bonds they were taking out this "insurance" on. Got that? Maybe not. It's so counter-intuitive it begs for repetition and illustration.
Credit default swaps did indeed start out as a kind of "insurance" against bad investments. AIG is, after all, an insurance company, right? So say you bought a bond. You were worried the bond might default rather than pay you back when it came due. You could hedge this risk by buying a credit default swap from a company like AIG. AIG would then, for a small fee, promise to pay you should your bond default. But they didn't stop there. And here's where it gets a wee bit crazy.
Imagine, if you will, your neighbor taking out a fire insurance policy on YOUR house. He only gets paid if YOUR house burns down. Now his only real financial stake in your house is hoping for some future fire. You might find that a wee bit odd. But the credit default swap market apparently did not, because that's exactly how it worked. Leave your house aside and say your neighbor thought your bond just might default. He didn't own the bond, but he wanted to make a bet that it was going to default. AIG would be happy to sell him his own credit default swap against the same bond - even though he didn't own the actual bond he was "insuring."
And it wouldn't stop with your neighbor. Anyone who wanted a piece of the action could buy into the credit default swap game. In effect it became a huge "gambling" operation, with credit default swap sellers banking on the overall success of the bond market, and credit default buyers banking on failure.
As you might imagine, this quickly got out of hand. Schieber quotes Gretchen Morgenson making the following observation:
When you look back with the benefit of hindsight, it is truly amazing how outsized A.I.G.’s insurance commitment [on credit default swaps alone] was, at $440 billion. After all, in 2005, when A.I.G. put many of these swaps on its books, the market value of the entire company was around $200 billion.
That means the geniuses at A.I.G. who wrote the insurance were willing to bet more than double their company’s value that defaults would not become problematic.
This gets to the other big problem with credit default swaps which Schieber kindly spares us by brushing past save for the indirect allusion quoted above. There were no regulatory requirements for a company to have the capital to back up the swaps they sold. Got that? AIG sold more than twice its entire net worth in credit default swap commitments. But it might just as well have sold ten times that. A hundred times that. (Alright, other less direct accounting practices constrained this to some extent, but no regulatory oversight of credit default swaps themselves did so.)
The point is, should the bond market ever really turn sour - and has it ever done so lately - AIG didn't have anything close to the capital it would need to honor its "insurance" commitments. That's exactly why more and more money is being funnelled through AIG in the name of "bailouts."
Basically, in bailing out AIG, US taxpayers are bailing out the glorified gambling debts of a financial system run amock. It goes FAR beyond bailing out merely those who had valid concerns about protecting their bond investments.
Just another great use for your tax dollars since you surely have nothing better to do with them.

The point you don't make directly about AIG, but sort of do circuitously, is that AIG being "too big to fail" took counterparty risk off the table for everyone that bought swaps from them.
The swaps I dealt with were held by a bankruptcy remote SPE as part of a syhthetic mortgage-backed CDO (another lovely topic). The trust was the seller of the CDSs, and the counterparties knew that if and when the trust ran out of money, there was no more "coverage". They were able to make their hedges or bets or whatever accordingly.
With AIG, I would guess that all of their counterparties knew that they would not be allowed to fail, and thus everyone played along trusting that the "coverage" would hold.
The instruments themselves are not the problem. That an institution would be allowed to issue them in great excess to what they might be able to pay, as you describe, certainly is a problem. I am very curious about he sort of risk managment and monitoring AIG had that let them get into that position. Did they have quants telling them, "it will never be a problem", or did they just not know or undertand how much they were on the hook for.
Here's a thought exercise I haven't completed....what if the part of AIG that sold the CDSs was allowed to fail, and thus not make good? If the counterparty was simply speculating, it would be like winning a bet with a guy that didn't pay up- it stinks, but you really weren't counting on the money anyway.
If the counterparty was actually hedging risk, then they would have to take the losses on the bonds, which would be an economic problem for that firm, and perhaps broadly- i.e. instead of AIG failing, now AIG has failed and CDS buyer-Bank X is also in trouble.
Good topic and good post. Thanks
and that you don't even examine whether a company of AIG's corporate portfolio/family is regulated -- I think the NY Department of Insurance would disagree with you, as would the NAIC
and you don't examine whether companies like AIG are required to have surplus to offset exposures on insuring bond risks
or are examined regularly by regulators to see whether their investment portfolio is excessively risky
no, everyone just assumes this is all just a big "error" on AIG's part.
I guess Hank Greenberg has fans everywhere. after all, why would anyone want to ask how a highly diversified international insurance conglomerate, with some of the industry's most savvy people working for it, could ever fail to understand the volatility of Credit Default Swaps. I mean, it's not like AIG had a financial products division, or had any experience whatever rating risks and observing its claims history.
a "mistake" by AIG?
really?
wow.
You might find this post helpful in your puzzlement at the discussion here so far. It's not something I've seen anyone distill into a single article yet, so please forgive me if you were expecting a more comprehensive treatment of the issue.