AIG-FP's 10-Q: Collateral was the Contagion

AIG's 10-Q for 3Q-2008 has some interesting information. (The 10-K was less helpful as so much disappeared into the maw of Leviathan. And, yes, I'm behind on my reading!) Looking at FAS-157's so-called "fair value" or "market valuation" (a.k.a. MTM) of super-senior CDS/CDO, as a percent of notional I see that the MTM for YTD 9-mos (i.e., excl. 2007 MTM losses) of "regulatory capital" relief transactions on corporate loans and prime RMBS the losses were 0% on $248.3bln notional. But on one deal of $1.6 bln notional, deemed no-longer-regulatory-relief, it was 25%. For AIG-FP's "arbitrage" multi-sector CDOs the MTM loss was a whopping $19.9 bln or 66% of $71.6 notional -- again at super-senior attachments. But for corporate CDO & CLO only 3% of notional, and 3% again on mezz' tranche reg' relief CDS -- mezz'! Total MTM for the 9-mos was $21.7 billion on $377.3 bln notional, or 6%. I can't find any loss or claims-paid data, except for 2a-7 and other ratings or valuation puts.

Later in the 'Q they do a two-scenario roll rate analysis. Roll rates 30-to-loss were 80% for 2006 & '07 vintages, 70% for '05, and 60% '04 & earlier; severities ranged from 50% to 60%. (i.e., ballpark-ish to others' estimates) Their NPV of losses were $7.8 & $12.0 billion in their scenarios A & B, respectively. This compares to a "FV" or MTM loss of $30.2 bln (cumulative, not YTD) and a collateralization burden of $32.8bln. Scenario B simply increased the roll rates & severities by 20% (but not exceeding 100%, of course).

So: AIG was trying to collateralize at 109% of "FV", at 421% of expected loss in scenario A, and at 273% of stressed loss in scenario B! Now imagine Citi, Merrill, UBS, and others also out there trying to collateralize to 109%. Maybe they started 109% of expected loss but with each twist of the spiral it went up & up, to 200% and beyond. ('Loss-actual' was rising at the same time, too, injecting some nitrous-oxide into the turbocharger.) In other words ... Collateral was the contagion!!!

And remember, Leviathan, d/b/a Maiden Lane II LLC, doesn't play the"FV" / MTM game. The Fed is a PV shop, but they don't disclose how they PV or their assumptions. If they recognize they can't run their bank on a MTM basis why can't they recognize the banks can't be regulated on a MTM basis? Like Ben Graham's aphorsim, solvency regulation of financial institutions is a "weighing" machine not a "voting" machine. MTM is a discipline, not a religion. (Like I said before, FASB has made it into a fatwa.)

JRB

4/1/09