Continued Harping on AIG Old News Shows Further Investigation Needed

Sorry for two AIG stories in succession; I just posted the other one when I saw the Bloomberg story just released, “Secret AIG Document Shows Goldman Sachs Minted Most Toxic CDOs.”

I’m a bit puzzled as to what the fracas is about here. The formerly redacted Schedule A, the list of the CDOs that the Fed purchased from various dealers through a vehicle called Maiden Lane III, has been out for nearly a month.

Tom Adams, assisted by Richard Smith, Andrew Dittmer, and me put together a detailed analysis of the transactions, prior to the release by Rep. Issa of the Congressional Oversight Panel of Schedule A, that showed much of the information was already public (see here, here, and here). With the full schedule available, we were able to remove some names from the original list (some of the transactions in the relevant portfolio clearly had not gone over to Maiden Lane III, so were were able to exclude them). The resulting spreadsheet included details like the transaction name, CUSIP, date of issuance of the CDO, original par amount, ratings at various points in time, lead bank, CDS counterparty, CDO manager.

The point of this exercise was to show that the Fed’s secrecy claim was bogus. The argument was that it needed to keep this information secret to protect Maiden Lane III’s positions from “traders” to maximize value. But if non-traders, with access to neither a Bloomberg data service or specialized databases that show CDO collateral, could put this much together from public information, clearly traders could go even further. Thus it was clear that it was the public, not the trading community, who was the real object of the secrecy campaign.

Moreover, the key point in the Bloomberg story tonight, that Goldman had the best perspective on what was going on in this portfolio at AIG, both through the CDOs it had insured with AIG and the CDOs it had created but other banks had insured with AIG, has also been in the public domain for nearly a month.

The problem is that, as bad as Goldman’s role here looks, the firm has a ready defense. Yes, it concentrated its CDOs at AIG, while other investment banks used the monolines much more actively (Goldman did use monolines, particularly ACA, but it appears only after AIG started to withdraw from the market). But AIG aggressively marketed its insurance as being “better” precisely because AIG posted collateral, the very feature that brought the giant insurer to ruin. By contrast, when CDOs insured by other banks cratered (mezz CDOs, which were composed heavily of BBB subprime bond tranches, as a product were almost without an exception a total wipeout), the party holding the AAA tranche faces considerable delay in collecting a payment (often, the payment is not due until the termination date of the underlying legal entity, which is often 40 years out). So why should Goldman buy an inferior product?

But interestingly, Goldman has not offered this seemingly obvious defense, which suggests that other reasons for preferring AIG may have been in play. It may have been that Goldman recognized that the monolines were not money good on the guarantees they were writing, but did not make as critical an examination of AIG (ie, Goldman goofed, but was able to play its way out of a weak position through its better view of what was happening and perhaps by clever use of its friends in high places), or that it saw AIG as particularly clueless and figured it was the easiest prey.

Or Goldman may simply be keeping mum on AIG generally to forestall investigation of its Abacus program, in which it created synthetic CDOs for the sole purpose of enabling the firm to put on a short position, putting its interest in direct conflict with the customers who took the long position in these trades. Note those deals were NOT included in Schedule A, but were part of the original AIG multi-sector CDO portfolio (ie, they still sit at AIG, rather than having been sold to Maiden Lane III).

So a further investigation of Goldman, SocGen, Deutsche Bank, the major AIG counterparties, is long overdue. Having wrestled with the data in the public domain, we can only conclude it is impossible at this juncture to understand the motives of the major players, which is essential to determining how much of this disaster was due to incompetence versus nefarious intent.

Now some may complain that people will continue to harp about AIG regardless, but that’s a spurious argument. Its rescue was a huge taxpayer commitment, done in great haste, and much of the background and the critical decisions remain in the dark. Sus looking details, like the decision to pay the counterparties at 100% of notional value of the CDS, were investigated by SIGTARP and criticized roundly in its report.

Note to Fed: if you want the media to stop worrying at the AIG bone, disclose all information that will give insight into the roles and objectives of the major AIG counterparties, including details of the Abacus deals. And if this step does not provide answers, it’s time for SIGTARP and the Financial Crisis Inquiry Commission to demand information from the counterparties themselves.