Wednesday, July 15, 2009

Megan Mcardle rips Matt Taibbi's Rolling Stone piece on Goldman Sachs

To give you a flavor of what I mean, Taibbi rants about how we knew derivatives were bad bad BAD! because they'd gone so badly wrong before:
There was only one problem with the deals: All of the wheeling and dealing represented exactly the kind of dangerous speculation that federal regulators are supposed to rein in. Derivatives like CDOs and credit swaps had already caused a series of serious financial calamities: Procter & Gamble and Gibson Greetings both lost fortunes, and Orange County, California, was forced to default in 1994. A report that year by the Government Accountability Office recommended that such financial instruments be tightly regulated - and in 1998, the head of the Commodity Futures Trading Commission, a woman named Brooksley Born, agreed. That May, she circulated a letter to business leaders and the Clinton administration suggesting that banks be required to provide greater disclosure in derivatives trades, and maintain reserves to cushion against losses.
But it's not clear how much derivatives regulation would have helped any of these three companies.  Gibson was defrauded by its bankers.  P&G wasn't; they spent a great deal of money unwinding their positions when the Treasurer realized they had a lot of exposure on a bad bet on falling interest rates.  Orange County, too, was making a massive, levered bet on a steep yield curve (roughly, a large difference between short and long term interest rates) that came undone when the yield curve flattened and interest rates rose.  Moderately complex derivatives allowed its idiot financial manager to take somewhat larger bets, but you can take massive, money losing bets without them.  At any rate, none of these derivatives have much to do with CDOs or CDSs; you might as well conflate stocks and bonds because they're both "securities".   No one, as far as I know, is now proposing that we need to curtail the use of interest rate swaps.

Did individual portfolio managers take on too much risk?  Yes.  Did some morons get sold these bonds without understanding what was going on?  Undoubtedly.  But Goldman's customers for CDOs are not little grannies who think a bond coupon is what you use to buy denture glue.  They're institutions who could reasonably be expected to understand the risks.  Which is why it is not, as Taibbi absurdly claims, "securities fraud" for Goldman to sell people mortgage-backed CDOs when they themselves were moderately short the overall housing market.

"That's how audacious these assholes are," says one hedge-fund manager. "At least with other banks, you could say that they were just dumb - they believed what they were selling, and it blew them up. Goldman knew what it was doing." I ask the manager how it could be that selling something to customers that you're actually betting against - particularly when you know more about the weaknesses of those products than the customer - doesn't amount to securities fraud.

"It's exactly securities fraud," he says. "It's the heart of securities fraud."

First of all, of course banks sell people positions they aren't themselves taking.  Sometimes the bank is right, and sometimes the customers are; differences of opinion are what make marriages and horse races.  Second of all, the banks that went down, the ones that arguably caused the financial crisis, were long their own toxic waste (and that of others).  Third of all, Goldman itself might argue that its mortgages were not as toxic as others, and for all I or Matt Taibbi know, they might be telling the truth.  Fourth of all, the disconnect between the underwriting and the customer side of the investment houses was not only legal, but in some cases, mandatory.  Excessive entanglement between the two is why Henry Blodget, whom Taibbi references elsewhere, has been banned from the securities industry for life.  That Taibbi could even ask how this was not securities fraud is really troubling.

Well worth reading in full.