Did you know that AIG has a Blog Relations department? For real. They sent out a big email earlier today, which wound up in places as varied as Dealbreaker and Welt Online, taking issue with the WSJ’s story about them this morning. Unfortunately, they seem incapable of writing in English, and even if you do try to decipher what they’re saying, it seems to be little more than "this isn’t news, it was on page 117 of our 10-Q".
The notional amount attributable to the cash settlement portion of the AIG Financial Products multi-sector credit default swap portfolio has been consistently included in the total AIG Financial Products multi-sector credit default swap exposure in AIG’s SEC filings and is explained on page 117 of AIG’s Quarterly Report on Form 10-Q for the period ending September 30, 2008.
I did end up talking to a human at AIG about this, since I wondered whether the company actually had any substantive issues with the WSJ story. It turns out that there is one big issue — that the amount of money in question is not $10 billion, as the Journal would have it, but…
…and there AIG goes quiet. They’re happy to tell you that the $10 billion is a notional amount, and not a mark-to-market loss: it’s AIG’s maximum possible loss, and the insurer does not at this point "owe Wall Street’s biggest firms about $10 billion", as the Journal says. But AIG won’t tell us how much it does owe on this book, so it’s impossible to tell whether its actual mark-to-market loss is close to $10 billion or not.
AIG is not disputing the main thrust of the story, which is that AIG Financial Products was stepping far beyond its remit as part of an insurance company. Most of the credit default swaps written by AIG were real insurance: they were sold to banks who held the securities in question and wanted to hedge their exposure.
But this $10 billion book wasn’t insurance at all, it was outright speculation. And now the US government is having to put up billions of dollars in collateral against those bets — bets which have gone very sour indeed.
I’m calling this one for the Journal. There might be a few minor errors when it comes to the specifics, I don’t know. But the big picture is clear. AIG insured banks, and it was necessary to bail out AIG in order to prevent a much larger domino effect caused by those banks not being paid out by their insurer. At the same time, however, there’s no reason to bail out the bits of AIG which were simply making speculative bets on the credit markets — and indeed there’s no reason why a triple-A insurer such as AIG should ever be making such speculative bets in the first place.
How much have those speculative bets cost? AIG won’t say. This time last month, it held a quarterly conference call, and in the slides for that call, on page 15, it gave some numbers for its total CDS exposure — both insurance and speculative. AIG wrote $71.6 billion of protection on multi-sector CDOs, and its mark-to-market loss, as of September 30, was $30.2 billion. That’s all we know so far. In October, that loss surely went up substantially; the loss on the speculative CDS might be much greater, in percentage terms, than the loss on the insurance CDS. But it’s probably safe to say that at a minimum, AIG’s mark-to-market losses on its speculative bets — losses for which the government has to provide collateral — are at least $4 billion or so.
That’s a lot of taxpayer money to put to use bailing out an insurance company’s prop desk. And given AIG’s recalcitrance in coming up with hard numbers which might contradict the Journal’s reporting, I do wonder why the Blog Relations department was so keen to send us all this note. They might have been better off just keeping quiet, frankly — especially when their formal statement is so incredibly stilted.