By James Kwak
ProPublica has a long and detailed story of Magnetar, the hedge fund that helped fuel the subprime bubble by providing the equity for new subprime collateralized debt obligations — precisely so that it could then go and short the higher-rated tranches. In other words, Magnetar wanted to short some really, really toxic CDOs. But either there weren’t enough toxic CDOs to short, or they weren’t toxic enough. So they provided the equity necessary to manufacture more toxic CDOs. Then they shorted them. Yes, the math works out.
Yves Smith told the story of Magnetar in her book ECONned
The Baseline Scenario 
JPMorgan Gets Into the Game — And Loses
You have GOT to read Magnetar’s response letter to ProPublica. In it, they say that they didn’t encourage the inclusion of more toxic mortgages in the CDOs. Rather, they didn’t care at all what was in the CDOs! From their letter:
“Magnetar employed no fundamental analysts to make its investments in CDOs.”
They didn’t even look to see what they were buying or shorting. They had their statistical models, and they figured that was enough.
I wonder if Magnetar thought to hedge against counterparty risk of their CDS? AIG failure to deliver collateral would have hurt Goldman, but could have destroyed smaller players like Magnetar and Burry’s hedge fund described in The Big Short.
Hi, it’s the first time I commented here. Here is a question that always puzzled me: I am a math major, and I don’t understand the phrase “the math works out”. I got the feeling that there is no way one can buy something and short it at the same time, then make money off it. The only way this would have worked is to mislead other people into bad decisions.
The Equity is a small but necessary part of the CDO. It is generally the most difficult part of the deal because it is the most exposed to loss. Maganetar’s shorting strategy meant that they expected to lose all of the equity that they were putting up. This is counter-intuitive as Equity investors almost always are betting on high ROI .
The math works something like this: Magnetar puts up $20 million for the equity part of a $1 billion deal (other CDO equity investors might be enticed to buy some equity also). Magnetar buys shorts on $X hundred million of the deal. As the CDO loses value, the equity loses and the short position gains. Magnetar is betting that the fall in value will be so great that their short position makes more than the loss of the equity.
The shorting here is buying credit default swaps. If you put $20 million into an equity tranche, one is able to buy say $100 million in credit default swaps. A failure brings in $100 million against a loss of $20 million.
Magnetar bought equity tranches where the cash flow was not sequestered based on pricing decline. They received cash flows of say 10 % for all tranches. On the other hand CDS “insurance” was priced very low. As an example, the front fee could be 1 % of coverage against 4 % a year in fees. Thus, most , if not all ,of the cash flows preceding payoff due to failure of the CDO would be covered from the CDO monthly cash flow.
The math works out especially if the CDO percentage of ownership compared to CDS coverage is very low.
Magnetar’s people understood the simplcity and the need for repeated transactions.
Along the way, Magnetar could have sold the CDS coverage and locked in the gain once they received more than they could lose on any equity tranche they bought.
Even then, they could hold the CDO for liquidation or run off cash flows.
Who was selling the CDOs?
JerryJ,
Who was selling the CDOs?
The banks created these CDO’s specifically to sell the equity tranche to Magnetar. The other tranches were peddaled into the world market. What was left was retained by the bank. According to the linked detailed article, JPM took a very hit on their holdings of Magnetar instigated deals.
John Paulson did the same thing. He did a ton of business with Deutsche Bank. One trader inside DB made DB a ton of money . The other traders thought he was bonkers.
The bulk of the people on Wall Street thought Magnetar was crazy. There were articles about how exposed Magnetar was.
What gets buried in the individual scamming accounts is that virtually everyone in big finance was convinced that the quants had designed products that were riskless. Even casual remarks that the quants were wrong brought on personal attacks. We are suffering the result of personal belief systems being inflicted on the world. Just because a person believes something does not turn it into a fact.
Very helpful. The question remains whether the quants believed because they had every incentive to believe (or at least, not to think too hard about the questions).
The truth is the math did not work. It only works for one party. Win/Lose. The structure of the deal should have had more equity to protect the debt. Then the equity investment they made would not have made sense.
http://en.wikipedia.org/wiki/There's_a_sucker_born_every_minute
ok, I’m gonna withhold what I think about a woman (said in the tone of Jack Nicholson’s Joker in revenge to Grissom in 1989’s Batman) who can write about the minutia of finance, but cannot correct 3 links in a post after 5 warnings from a reader. Her initials are Y. S. …………A woman!!!!!!!!!!
ok, this GD link doesn’t work. try this. Think of it as therapy…or Spanish lessons, or spanish lessons with therapy. Shpanipy
the structure of the deal was unworkable, because it was structured to “help” people buy houses at the top of the bubble. If you bought a house with 100% equity at the top of the bubble, you still lost money. Magnetar thought we were at the top of a bubble. Their counterparties were too stupid to imagine a world in which house prices could actually go down.
Thanks Jerry :)
From my understanding, ProPublica does all the heavy lifting for this story and This American Life (Public Radio International) and the Planet Monkey Brains at NPR take the credit? Is that about right? Lamestream media, including public radio, has been a hapless, hopeless follower of news related to the financial meltdown. They still don’t understand the housing bubble, and they continue to booster news that isn’t completely dire, as if the U.S. economy is about to turn some corner.
The heaviest lifting ProPublica did was hire another firm to survey Magnetar’s deals and determine their rate of failure relative to the rest of the CDO market. If you listen to the TAL story, it’s pretty clear that Alex is doing interviews alongside the two PP reporters.
I think I know how Magnetar avoided CDS counterparty risk. Many of their CDOs were synthetic, and probably contained collateral composed not of MBS, but of the long side of CDS written on MBS. So picture this. Magnetar puts together a CDO based on crappy MBS, and wants to short it but can’t since no one will take the long side of the deal. So instead, they buy CDS on the individual MBS bonds that collateralize the original CDO, and get a CDO manager to package up the long side of those trades into a new synthetic CDO. If the first CDO fails, the second one will too; but Magnetar doesn’t care because it will make so much money on the short position on the first CDO. And if they by some miracle don’t fail, Magnetar has TWO equity tranches returning a hefty percentage, a fraction of which is really just Magnetar paying itself.
Felix Salmon compares regular and synthetic CDOs.
Should ‘analysts’ be ‘analysis’?