Illegal or not, Goldman's maneuvers betrayed its own principles

By Allan Sloan
Tuesday, April 20, 2010

"Our assets are our people, capital and reputation. If any of these is ever diminished, the last is most difficult to restore."

-- From Goldman Sachs

Business Principle No. 2

The hot question on Wall Street these days is whether Goldman Sachs violated securities laws in the Abacus case, as the Securities and Exchange Commission charges. But I don't much care about the legality of what Goldman allegedly did, because something doesn't have to be illegal to be wrong. And almost everything about the Abacus 2007-AC1 "synthetic collateralized debt obligation" deal was wrong.

No, not for legal reasons, but for the reason Goldman cited in its Principle No. 2. The firm not only put its reputation at risk but also, by its own account, managed to lose money. It wasn't just dumb, it was double dumb.

Let me explain. Abacus, as we'll soon see, wasn't an investment vehicle -- it was a gambling vehicle. By definition, with a synthetic CDO, someone's gain is someone else's loss. Think of it as a big-bucks poker game in which Goldman lined up three high rollers, provided them a private room, helped negotiate the rules and took a fat fee for bringing everyone together.

When you host one of these games, you not only have to make sure that it isn't rigged, you also can't afford to have it appear rigged, in case the losers (or the people who cover their tab) go to the cops. In addition, if you're smart -- what Goldman calls "long-term greedy" -- you make sure the game doesn't help undermine the society that makes your business possible.

Goldman, based on what I think I know about this case, violated both of these principles. (The firm, not surprisingly, declined to comment beyond what it has already said.)

The key to grasping the Abacus deal is something I didn't see in the SEC's records. It's that a synthetic CDO, such as Abacus 2007-AC1, involves betting, not investing. In this case, it allowed players to bet on the direction in which a "reference portfolio" of securities based on junk mortgages would move.

All the Abacus players knew there was someone betting against them. It's not the same thing that I wrote about in 2007, when I showed how Goldman had peddled a security, GSAMP 2006-S3, based on mega-trashy second mortgages, then later bet that such mortgages would decline in value, but no word about this to its clients.

In the Abacus case, Paulson & Co., then a little-known operation, bet that the reference portfolio would tank, which it did almost immediately -- hence the bad smell to this deal. Three players bet that it wouldn't: ACA, a firm that puts its name on the deal; a German bank called IKB; and Goldman itself. ACA lost $840 million, which ultimately came out of the pockets of British taxpayers because Royal Bank of Scotland, which got a government bailout, had guaranteed ACA's ability to cover its bet. IKN, which also got a government bailout, lost $150 million, which ultimately came out of the pockets of German taxpayers, and Goldman lost $90.9 million, which more than wiped out its $15 million fee for setting up the deal. Goldman seems to have taken the least risky piece of the deal, but got bagged anyway.

CONTINUED     1        >

© 2010 The Washington Post Company