A Few Brief (But Hopefully Deep) Thoughts

John Authers at the Financial Times has an excellent article out about synthetic CDOs.  I’ve been discussing this with some folks, and we seem to be on the same page.  I think there’s a serious question about the behavior of the parties involved on all sides of these transactions.  While I think ultimate culpability lies beyond the agents involved, I think it’s safe to say we have to have better, clearer lines of demarcation about what’s acceptable and what isn’t.  I won’t get into more depth here, that’s what a book is for.

About the Goldman emails: that seems like there’s a lot more smoke than fire.  I don’t begrudge Goldman for making the trades they did, when they reached that point of recognition in December ’06.  In fact, if anyone bothered to ask me while I was at Wachovia in their corporate treasury, I would’ve begged to offer that we needed to take many of the same actions.  Would I have recommended going the same size as Goldman did?  Probably not, but it would’ve helped the bank avoid being embroiled in this with Citi and Wells Fargo later:

But alas, I was already gone from the bank and they decided to purchase Golden West.  Or as I refer to it: Money Store, Part Deux.  Kind of like “Hot Shots Part Deux!” except much more overvalued…

When is a hedge not a hedge?  When it becomes a directional trade, of course.  And the only difference between the two at that point is the size of the exposure.  Would I have advocated legging into a long position on ABX and CMBX protection?  Absolutely.  Would I have also advocated winding down and selling the loans we had, in spite of the losses on the balance sheet?  Yes, because it’s like dealing with Band-Aids.  Do you try to p e e l it off s  l  o  w  l  y or just rip it off?  Rip it off, of course.

That is all…

On Goldman, Synthetic Longs & Shorts, and S_ _ _ Sandwiches

The news about Goldman has definitely been making its way around.  Just look here, here, here, here and here.  There’s something particularly nostalgic and ‘newspaperish’ in the Huffington Post’s presentation: you’d think it was V-E Day all over again.

For those who want to see the pitchbook, it’s included here:

View this document on Scribd

This slideshow does a good job of explaining the transaction…

But I want to take a different angle on this story.  I want to look at it from the decision-making process that Goldman the firm was faced with.  Not Goldman the marketing machine, not the Goldman you see in its company history pages on its website, but the Goldman that has to manage risk.  What were its options?  Why did it choose the option it chose?  I’m not going to talk about the legality/illegality of the deal, because I’m not a lawyer.  But the characteristics of the market at that time were such that were plenty of bad decisions made.  Some on the part of the structurers, but also a bunch by the investors, who supposedly had a level of “sophistication.”

The Trade

We all know John Paulson was looking to make a bet against the housing market.  So via ACA, he created a CDO of the MBS he wanted to short.  Goldman, as Paulson’s broker, facilitated his request by selling him the credit protection against the securities that were put in the CDO.  Essentially, Goldman created a bespoke credit default swap index in handling Paulson’s end of the trade.

I emphasized bespoke because it should tell you something about the liquidity risk inherent in such a trade.  There’s tons of it. There are liquid credit indices, which you can find at Markit’s site.  Liquid credit indices are just like liquid stocks.  You can buy/sell them in size, and pay very little for the privilege.  Illiquid credit protection, however, is a whole other ballgame.  You can’t get in/out of positions easily even in good times and you’re going to pay through the nose to do it.  And if you try to do a trade in size?  Puhleeeze…

This also illustrates why it’s laughable to think CDS have anything to do with making the sovereign credit situation worse.  These are synthetic shorts when you buy them.  If they didn’t exist, there would only be one option: sell the cash bonds and watch the yields & spreads blow out even higher than where they are.

Goldman’s Decisions

But in the process of selling Paulson the protection he wanted, Goldman took on a position itself as well: selling credit protection is a synthetic long on the credit risk of the position.  It’s not like Goldman lent ACA or the SPEs that created those subprime MBS the money to originate the loans, but in selling the protection, you’re expressing a view on the credit risk that’s there – a view that says it’s overpriced to you.  This is actually a cheap way to gain credit exposure to an MBS, a company, you name it.  I say it’s cheap because you’re gaining credit exposure by betting that the entity won’t default on its obligations and you’re not putting actual balance sheet at risk via traditional lending.

So the first decision you’re confronted with: do you want to keep this long credit exposure?  If you answer ‘Yes’ then you do nothing else.  But if you don’t want to keep it, you have to figure out a way to off-load it somehow.  There are several ways you can short/get rid of this risk:

• Buy credit protection yourself and create a synthetic short position in the process.
• You can sell the assets off your balance sheet.
• Or you can create a synthetic loan sale in the form of a CDO if you don’t have the assets on your books, but still have credit risk in the form of credit protection you sold.

Buying protection on an illiquid index will be expensive.  You may not want to take the risk that buying the protection from Credit Suisse, JP Morgan, or some other firm will cost more than what you charged Paulson for it.  You may have hedged the risk, but you have a P&L issue as a result.  These assets were not on Goldman’s balance sheet, so no bulk loan/securities sale would work.  You’re really left with exiting the position by creating a CDO.  Bringing the CDO to market also has the benefit of generating a positive contribution to the P&L, even if it’s only a modest contribution.  The real goal is to off-load the synthetic long position you created by selling the CDS to Paulson in the first place.

Final Points

I’m not a Goldman apologist.  I’m pretty sure they made some bad decisions in this situation, namely getting too chummy with a client to get additional deals/trades to work for them.  The disclosure issue?  Is it a case of withholding information or if they did disclose it, would that have been a case of releasing MNPI?  I don’t know, but it seems to me the case is far from a slam dunk for either side.

As for AIG?  I still remember this quote from Andrew Ross Sorkin’s “Too Big To Fail” from Sam Cassano at AIG’s Financial Products unit:

My brother works at Goldman, and he’s an idiot!

Ah, Sam… Really?  If that’s true, that pretty much makes you full-retard.  $6bn in wrap guarantees?  What were you guys thinking?

So folks, next time someone hands you a sandwich, you might want to check what’s between those pieces of bread…