News for hackers. Stuff that matters.
Re:U.S. Set to Sue a Dozen Big Banks Over Mortgages by yummyfajitas on Friday September 02, 2011 @05:49AM (#2953531)
No.

Please spend some time learning what a CDO is, and the reason why will be fairly clear.

http://en.wikipedia.org/wiki/Collateralized_debt_obligation#...

The junior and senior tranches can only be equally risky if the probability distribution is perfectly binary (i.e., the only possibilities are either all borrowers pay 100% back, or all borrowers pay 0% back). This is simply a mathematical identity.

Re:U.S. Set to Sue a Dozen Big Banks Over Mortgages by arethuza on Friday September 02, 2011 @06:00AM (#2953550)
Fair enough - I wouldn't claim to be an expert!

So what did go wrong with the way CDOs were supposed to work and they did work in reality?

Edit:

Actually - genuine question here. When someone says "This is simply a mathematical identity." - how is all of this stuff grounded in reality - do people actually do controlled experiments to work out if any of this stuff is scientific?

Are any of these theories falsifiable?

Re:U.S. Set to Sue a Dozen Big Banks Over Mortgages by yummyfajitas on Friday September 02, 2011 @08:41AM (#2953864)
So what did go wrong with the way CDOs were supposed to work and they did work in reality?

Counterparty risk.

Home loan CDO's were supposed to allow people to take a long position on housing. Assuming counterparties didn't go belly up, they would have done exactly that - the owners of various securities would have made money when housing went up and lost when housing went down.

When someone says "This is simply a mathematical identity." - how is all of this stuff grounded in reality - do people actually do controlled experiments to work out if any of this stuff is scientific?

It's just basic arithmetic. The borrowers are going to pay out a certain amount of money. The senior tranch gets paid first. If any money is left, the junior tranch gets paid. Short of politicians overriding the rule of law (e.g. the GM bankruptcy), there is no scenario where the junior tranch gets a payout and the senior tranch does not.

Re:U.S. Set to Sue a Dozen Big Banks Over Mortgages by arethuza on Friday September 02, 2011 @08:58AM (#2953912)
"Counterparty risk."

Presumably you mean the people taking out the mortgages?

Re:U.S. Set to Sue a Dozen Big Banks Over Mortgages by yummyfajitas on Friday September 02, 2011 @09:39AM (#2954036)
The counterparties that pose problems are the people who issued the CDO or insured it - think AIG, BankAm or Lehman.

The possibility of borrowers defaulting is already incorporated into the CDO.

Re:U.S. Set to Sue a Dozen Big Banks Over Mortgages by Klinky on Friday September 02, 2011 @07:32AM (#2953711)
The junior and senior tranches can only be equally risky if the probability distribution is perfectly binary (i.e., the only possibilities are either all borrowers pay 100% back, or all borrowers pay 0% back). This is simply a mathematical identity.

Or if you're in bed with ratings agencies who will rate whatever, however you want & you have an entire industry that will take their word for it.