AIG's Not Very Transparent List of Counterparties

It's good that AIG has released a list of its counterparties. But if it really believes in "the importance of upholding a high degree of transparency with respect to the use of public funds", this is a very odd way of releasing the information.

If you're not already familiar with the intricacies of AIG's operations, it's very easy to just start adding up the numbers in the various appendices, coming up with a kind of bailout league table: Goldman Got $12.9 billion! Barclays got $8.5 billion! But in fact it's much more complicated than that.

There are four appendices in all. Before we get to them, it's worth reading a bit of Gretchen Morgenson today:

Even A.I.G.'s own independent directors haven't been told which of the counterparties were paid...
Such secrecy raised hackles because the insurance claims were paid off in full, even though widespread defaults on the underlying debt have not occurred. Why, many people wonder, did the Fed make A.I.G.'s counterparties whole on losses that have not happened yet?

What Morgenson is talking about here is the second of the four appendices: the payments made by the company known as "Maiden Lane III". After banks insured their assets against default, AIG essentially used Maiden Lane III to take those assets onto its own books, thereby allowing it to cancel out the insurance contracts. The big winners here are SocGen and Goldman Sachs -- and it's worth noting that unlike the first appendix, where the counterparties are helpfully listed in order of size, in the second appendix there seems to be no particular order at all, and the two biggest recipients of government money are hidden in the middle of the list.

The other three appendices are not in the same class: they don't really constitute government giveaways in the same way. The first lists collateral postings which AIG has made but which haven't really been spent: pace Morgenson, if the losses never happen, then AIG gets all that money back. The third appendix is a list of states, which pretty obviously was included to make it seem as though public money was somehow just getting shunted around and returned to taxpayers in some other form.

The final appendix is a list of AIG's securities lending counterparties -- this is pretty much meaningless, and these counterparties hadn't bought any type of insurance from AIG at all. Instead they had simply borrowed securities from AIG, posting collateral of their own; when they returned the securities some time later, they got their collateral back. In the interim, AIG had managed to invest that collateral very badly, so it had to make up the losses itself. But those losses were not in any way related to the counterparties who borrowed AIG's securities, and it doesn't really make a lot of sense to think of those counterparties as being bailed out to the tune of $43.7 billion. If AIG had been liquidated, those counterparties would at the very least have kept hold of the securities they'd borrowed, instead of giving them back -- including the securities which had gone up in value rather than down. So the maximum loss to the counterparties would have been much smaller than $43.7 billion.

In any event, so many of the counterparties on this list had hedged their AIG exposure that it's massively oversimplifying matters to conclude that even the banks with the biggest exposures on the second appendix are the ones which effectively got the biggest government bailout. It's not nearly as simple as that -- and AIG should be much more upfront about such matters than it is being with this release.