Losses on the Super Senior positions contributed approximately three quarters of the CDO desk's total losses (or 50% of UBS's total losses) as at 31 December 2007...
Of the total USD 50 bn Super Seniors held by UBS, UBS purchased USD 20.8 bn of these Super Seniors from third parties.
UBS wasn't just holding onto its own super-senior tranches, it bought more than $20 billion from other banks as well! And it had some inventive ideas when it came to hedging the credit risk on these instruments, too. Consider the largest bucket of super-senior tranches:
Amplified Mortgage Portfolio ("AMPS") Super Seniors: these were Super Senior positions where the risk of loss was initially hedged through the purchase of protection on a proportion of the nominal position (typically between 2% and 4% though sometimes more). This level of hedging was based on statistical analyses of historical price movements that indicated that such protection was sufficient to protect UBS from any losses on the position. Much of the AMPS protection has now been exhausted, leaving UBS exposed to write-downs on losses to the extent they exceed the protection purchased. As at the end of 2007, losses on these trades contributed approximately 63% of total Super Senior losses.
I've read this quite a few times, now, and I still don't understand what it means. I think it means that UBS took out insurance against the value of the security falling between 2% and 4%: "the purchase of protection" here refers to protection against market moves, and not to CDS protection against an event of default. But I'm not sure. In any case, it was clearly a license to print money for the mortgage desk:
A hedging methodology enabled the desk to buy relatively low levels of market loss protection (generally 2 to 4% and sometimes more), and the desk considered the position as fully hedged.
It's the arbitrage that wasn't. You buy a security yielding more than your internal cost of funds, and which has a zero risk weighting, you "fully hedge" it, and you get a positive carry. What's not to love? At this point it's easy to see how UBS ended up with $50 billion of these things.
One normally hopes that a human being somewhere in most banks would look at a $50 billion exposure based only on "statistical analyses of historical price movements" and call a halt to what could be a very dangerous venture. Evidently no human at UBS did so.